OFFICE OF PORTFOLIO ANALYSIS September 5,...
Transcript of OFFICE OF PORTFOLIO ANALYSIS September 5,...
OFFICE OF PORTFOLIO ANALYSIS
September 5 2012
The Office of Portfolio Analysis (OPA)
OPA was established in 2011 to provide multiple services
Scientific portfolio analysishellip
o of all Common Fund initiatives
o in response to requests by senior leadership
Coordinate portfolio analysis activities across NIH
Train NIH staff to promote the effective use of analytical tools
o Regularly scheduled courses
o Ad hoc consultations
Improve portfolio analysis at NIH (see next slide)
Office of Portfolio Analysis (OPA) DPCPSI OD 2
Improve Portfolio Analysis at NIH
Solve NIH-wide limitations in cross-referencing databases
Build new computational tools that address specific NIH needs
Model NIH output and health impact accurately
o Advanced bibliometrics
Track progress in the new field of ldquoscience of sciencerdquo and find
useful methods and synergies among the parallel efforts inhellip
o Academia
o Other government agencies
o Private sector
Adopt new methods that meet NIH needs and provide training
opportunities for portfolio analysis stakeholders at NIH
Office of Portfolio Analysis (OPA) DPCPSI OD 3
Scientific Portfolio Analysis at NIH
Data-driven approaches to program development
What is the best way to accelerate scientific progress
An example
OPA analysis of 38 Metabolomics Centers in the US
most of which receive funding from NIH (~$70M in 2010)
Office of Portfolio Analysis (OPA) DPCPSI OD 4
$225M Global Investment in Metabolomics in FY10
bull Netherlands Metabolomics Center $67M
bull BBSRC UK Plant and Microbial Metabolomics $104M
bull Canadian Human Metabolome Database $81M
Office of Portfolio Analysis (OPA) DPCPSI OD 5
Europe
Japan
FY09 Metabolomics Co-authorship Networks
USA
Office of Portfolio Analysis (OPA) DPCPSI OD 6
Scientific Portfolio Analysis at NIH
What is the best way to accelerate scientific progress
OPA analysis of 38 Metabolomics Centers in the US most of which receive funding from NIH (~$70M in 2010)
These Centers have overlapping goals yet operated in isolation
with little coordination or collaboration
Portfolio analysis shaped a new Metabolomics RFA aimed at
improving coordination and leveraging existing resources
Office of Portfolio Analysis (OPA) DPCPSI OD 7
o httpdpcpsinihgovportfolio_analysis
Coordination of Portfolio Analysis Efforts at NIH
Portfolio Analysis Workshop (February 6 Natcher Conf Ctr)
Portfolio Analysis Symposium (July 23-24 Natcher Conf Ctr)
Build a computer lab to tailor existing and new computational tools to
NIH needs and to train NIH staff in their use
Centralized web-based repository to disseminate computational tools
Standing trans-NIH Working Group
Office of Portfolio Analysis (OPA) DPCPSI OD 8
Goals of the Portfolio Analysis Workshop Feb 5-6 2012
Discuss perceived needs in portfolio analysis
Use needs assessment to plan the Portfolio Analysis Symposium
(July 23-24 Natcher Conf Ctr)
Cover topics of broad interest to NIH decision-makers including
o Strategic planning
o Uses of portfolio analysis
o Overlap in NIH portfolios
o Measuring impact
o New portfolio analysis tools
o Identification of emerging areas
Office of Portfolio Analysis (OPA) DPCPSI OD 9
90
80
70
60
50
40
30
20
10
0
Outcomes of the Portfolio Analysis Workshop
o Full registration within 48 hrs of the announcement
o gt500 participants ndash approved for ESA training credit
Office of Portfolio Analysis (OPA) DPCPSI OD 10
Outcomes of the Portfolio Analysis Workshop
Survey results
Highest priority Topics included
o Measuring impact 47
o Gaps and overlap in NIH portfolios 30
o Identification of emerging areas 18
o Categorizing portfolios 5
What needs should OPA try to address
o Build better tools easier to use tools 65
o Provide training and support 50
o Develop targeted case studies 15
Office of Portfolio Analysis (OPA) DPCPSI OD 11
Goals of the Portfolio Analysis Symposium July 23-24 2012
Bring outside experts from academia government and the
private sector to discuss and demo state-of-the-art
approaches in scientific portfolio analysis
Choose those with expertise in areas identified as critical and
of greatest interest to portfolio analysis stakeholders at NIH
o Measuring Impact
o Identifying Gaps and Impact
o Identifying Emerging Areas
Facilitate the development of collaborations that address NIH
needs in scientific portfolio analysis
Office of Portfolio Analysis (OPA) DPCPSI OD 12
Outcomes of the Portfolio Analysis Symposium
Links to the Symposium agenda and archived videocast are
available on the OPA web site
httpdpcpsinihgovopaindexaspx
Outside experts presented new tools and approaches in each
topic one of these tools (Sci2) was particularly well received
by Symposium attendees and is being added to QVR
The Symposium resulted in several collaborations Examples
include OPA initiatives in building tools for co-author network
analysis analysis of patent development and tracking of
patent licenses
Office of Portfolio Analysis (OPA) DPCPSI OD 13
The Office of Portfolio Analysis (OPA)
OPA was established in 2011 to provide multiple services
Scientific portfolio analysishellip
o of all Common Fund initiatives
o in response to requests by senior leadership
Coordinate portfolio analysis activities across NIH
Train NIH staff to promote the effective use of analytical tools
o Regularly scheduled courses
o Ad hoc consultations
Improve portfolio analysis at NIH
Office of Portfolio Analysis (OPA) DPCPSI OD 14
Build Computational Tools and Train NIH Staff
Office of Portfolio Analysis (OPA) DPCPSI OD
OPA
TOOLS
LAB
12 laptops
Two workstations Building 1 Room B301
Smartboard
Linux server
Videoconferencing capabilities
15
OPA Consultations with NIH Staff
IC Contact Activity
NCCAM John Williamson Consult
NCI John Hewes Consult amp Collaboration
NCI Tanya Agurs-Collins Consult
NHLBI Marc Charette Consult amp Collaboration
NIAMS Faye Chen Consult
NIAMS Anita Linde Consult
NIDDK Lisa Spain Consult
NIGMS Michelle Hamlet Consult
NIGMS Ward Smith Consult
NIGMS Kelley Smith Consult amp Training
NIMH Yancy Bodenstein Consult amp Collaboration
NIMHD Ligia Artiles Consult amp Training
NLM Alan Vanbierlet Consult ampTraining
Office of Portfolio Analysis (OPA) DPCPSI OD 16
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 17
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
The Office of Portfolio Analysis (OPA)
OPA was established in 2011 to provide multiple services
Scientific portfolio analysishellip
o of all Common Fund initiatives
o in response to requests by senior leadership
Coordinate portfolio analysis activities across NIH
Train NIH staff to promote the effective use of analytical tools
o Regularly scheduled courses
o Ad hoc consultations
Improve portfolio analysis at NIH (see next slide)
Office of Portfolio Analysis (OPA) DPCPSI OD 2
Improve Portfolio Analysis at NIH
Solve NIH-wide limitations in cross-referencing databases
Build new computational tools that address specific NIH needs
Model NIH output and health impact accurately
o Advanced bibliometrics
Track progress in the new field of ldquoscience of sciencerdquo and find
useful methods and synergies among the parallel efforts inhellip
o Academia
o Other government agencies
o Private sector
Adopt new methods that meet NIH needs and provide training
opportunities for portfolio analysis stakeholders at NIH
Office of Portfolio Analysis (OPA) DPCPSI OD 3
Scientific Portfolio Analysis at NIH
Data-driven approaches to program development
What is the best way to accelerate scientific progress
An example
OPA analysis of 38 Metabolomics Centers in the US
most of which receive funding from NIH (~$70M in 2010)
Office of Portfolio Analysis (OPA) DPCPSI OD 4
$225M Global Investment in Metabolomics in FY10
bull Netherlands Metabolomics Center $67M
bull BBSRC UK Plant and Microbial Metabolomics $104M
bull Canadian Human Metabolome Database $81M
Office of Portfolio Analysis (OPA) DPCPSI OD 5
Europe
Japan
FY09 Metabolomics Co-authorship Networks
USA
Office of Portfolio Analysis (OPA) DPCPSI OD 6
Scientific Portfolio Analysis at NIH
What is the best way to accelerate scientific progress
OPA analysis of 38 Metabolomics Centers in the US most of which receive funding from NIH (~$70M in 2010)
These Centers have overlapping goals yet operated in isolation
with little coordination or collaboration
Portfolio analysis shaped a new Metabolomics RFA aimed at
improving coordination and leveraging existing resources
Office of Portfolio Analysis (OPA) DPCPSI OD 7
o httpdpcpsinihgovportfolio_analysis
Coordination of Portfolio Analysis Efforts at NIH
Portfolio Analysis Workshop (February 6 Natcher Conf Ctr)
Portfolio Analysis Symposium (July 23-24 Natcher Conf Ctr)
Build a computer lab to tailor existing and new computational tools to
NIH needs and to train NIH staff in their use
Centralized web-based repository to disseminate computational tools
Standing trans-NIH Working Group
Office of Portfolio Analysis (OPA) DPCPSI OD 8
Goals of the Portfolio Analysis Workshop Feb 5-6 2012
Discuss perceived needs in portfolio analysis
Use needs assessment to plan the Portfolio Analysis Symposium
(July 23-24 Natcher Conf Ctr)
Cover topics of broad interest to NIH decision-makers including
o Strategic planning
o Uses of portfolio analysis
o Overlap in NIH portfolios
o Measuring impact
o New portfolio analysis tools
o Identification of emerging areas
Office of Portfolio Analysis (OPA) DPCPSI OD 9
90
80
70
60
50
40
30
20
10
0
Outcomes of the Portfolio Analysis Workshop
o Full registration within 48 hrs of the announcement
o gt500 participants ndash approved for ESA training credit
Office of Portfolio Analysis (OPA) DPCPSI OD 10
Outcomes of the Portfolio Analysis Workshop
Survey results
Highest priority Topics included
o Measuring impact 47
o Gaps and overlap in NIH portfolios 30
o Identification of emerging areas 18
o Categorizing portfolios 5
What needs should OPA try to address
o Build better tools easier to use tools 65
o Provide training and support 50
o Develop targeted case studies 15
Office of Portfolio Analysis (OPA) DPCPSI OD 11
Goals of the Portfolio Analysis Symposium July 23-24 2012
Bring outside experts from academia government and the
private sector to discuss and demo state-of-the-art
approaches in scientific portfolio analysis
Choose those with expertise in areas identified as critical and
of greatest interest to portfolio analysis stakeholders at NIH
o Measuring Impact
o Identifying Gaps and Impact
o Identifying Emerging Areas
Facilitate the development of collaborations that address NIH
needs in scientific portfolio analysis
Office of Portfolio Analysis (OPA) DPCPSI OD 12
Outcomes of the Portfolio Analysis Symposium
Links to the Symposium agenda and archived videocast are
available on the OPA web site
httpdpcpsinihgovopaindexaspx
Outside experts presented new tools and approaches in each
topic one of these tools (Sci2) was particularly well received
by Symposium attendees and is being added to QVR
The Symposium resulted in several collaborations Examples
include OPA initiatives in building tools for co-author network
analysis analysis of patent development and tracking of
patent licenses
Office of Portfolio Analysis (OPA) DPCPSI OD 13
The Office of Portfolio Analysis (OPA)
OPA was established in 2011 to provide multiple services
Scientific portfolio analysishellip
o of all Common Fund initiatives
o in response to requests by senior leadership
Coordinate portfolio analysis activities across NIH
Train NIH staff to promote the effective use of analytical tools
o Regularly scheduled courses
o Ad hoc consultations
Improve portfolio analysis at NIH
Office of Portfolio Analysis (OPA) DPCPSI OD 14
Build Computational Tools and Train NIH Staff
Office of Portfolio Analysis (OPA) DPCPSI OD
OPA
TOOLS
LAB
12 laptops
Two workstations Building 1 Room B301
Smartboard
Linux server
Videoconferencing capabilities
15
OPA Consultations with NIH Staff
IC Contact Activity
NCCAM John Williamson Consult
NCI John Hewes Consult amp Collaboration
NCI Tanya Agurs-Collins Consult
NHLBI Marc Charette Consult amp Collaboration
NIAMS Faye Chen Consult
NIAMS Anita Linde Consult
NIDDK Lisa Spain Consult
NIGMS Michelle Hamlet Consult
NIGMS Ward Smith Consult
NIGMS Kelley Smith Consult amp Training
NIMH Yancy Bodenstein Consult amp Collaboration
NIMHD Ligia Artiles Consult amp Training
NLM Alan Vanbierlet Consult ampTraining
Office of Portfolio Analysis (OPA) DPCPSI OD 16
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 17
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
Solve NIH-wide limitations in cross-referencing databases
Build new computational tools that address specific NIH needs
Model NIH output and health impact accurately
o Advanced bibliometrics
Track progress in the new field of ldquoscience of sciencerdquo and find
useful methods and synergies among the parallel efforts inhellip
o Academia
o Other government agencies
o Private sector
Adopt new methods that meet NIH needs and provide training
opportunities for portfolio analysis stakeholders at NIH
Office of Portfolio Analysis (OPA) DPCPSI OD 3
Scientific Portfolio Analysis at NIH
Data-driven approaches to program development
What is the best way to accelerate scientific progress
An example
OPA analysis of 38 Metabolomics Centers in the US
most of which receive funding from NIH (~$70M in 2010)
Office of Portfolio Analysis (OPA) DPCPSI OD 4
$225M Global Investment in Metabolomics in FY10
bull Netherlands Metabolomics Center $67M
bull BBSRC UK Plant and Microbial Metabolomics $104M
bull Canadian Human Metabolome Database $81M
Office of Portfolio Analysis (OPA) DPCPSI OD 5
Europe
Japan
FY09 Metabolomics Co-authorship Networks
USA
Office of Portfolio Analysis (OPA) DPCPSI OD 6
Scientific Portfolio Analysis at NIH
What is the best way to accelerate scientific progress
OPA analysis of 38 Metabolomics Centers in the US most of which receive funding from NIH (~$70M in 2010)
These Centers have overlapping goals yet operated in isolation
with little coordination or collaboration
Portfolio analysis shaped a new Metabolomics RFA aimed at
improving coordination and leveraging existing resources
Office of Portfolio Analysis (OPA) DPCPSI OD 7
o httpdpcpsinihgovportfolio_analysis
Coordination of Portfolio Analysis Efforts at NIH
Portfolio Analysis Workshop (February 6 Natcher Conf Ctr)
Portfolio Analysis Symposium (July 23-24 Natcher Conf Ctr)
Build a computer lab to tailor existing and new computational tools to
NIH needs and to train NIH staff in their use
Centralized web-based repository to disseminate computational tools
Standing trans-NIH Working Group
Office of Portfolio Analysis (OPA) DPCPSI OD 8
Goals of the Portfolio Analysis Workshop Feb 5-6 2012
Discuss perceived needs in portfolio analysis
Use needs assessment to plan the Portfolio Analysis Symposium
(July 23-24 Natcher Conf Ctr)
Cover topics of broad interest to NIH decision-makers including
o Strategic planning
o Uses of portfolio analysis
o Overlap in NIH portfolios
o Measuring impact
o New portfolio analysis tools
o Identification of emerging areas
Office of Portfolio Analysis (OPA) DPCPSI OD 9
90
80
70
60
50
40
30
20
10
0
Outcomes of the Portfolio Analysis Workshop
o Full registration within 48 hrs of the announcement
o gt500 participants ndash approved for ESA training credit
Office of Portfolio Analysis (OPA) DPCPSI OD 10
Outcomes of the Portfolio Analysis Workshop
Survey results
Highest priority Topics included
o Measuring impact 47
o Gaps and overlap in NIH portfolios 30
o Identification of emerging areas 18
o Categorizing portfolios 5
What needs should OPA try to address
o Build better tools easier to use tools 65
o Provide training and support 50
o Develop targeted case studies 15
Office of Portfolio Analysis (OPA) DPCPSI OD 11
Goals of the Portfolio Analysis Symposium July 23-24 2012
Bring outside experts from academia government and the
private sector to discuss and demo state-of-the-art
approaches in scientific portfolio analysis
Choose those with expertise in areas identified as critical and
of greatest interest to portfolio analysis stakeholders at NIH
o Measuring Impact
o Identifying Gaps and Impact
o Identifying Emerging Areas
Facilitate the development of collaborations that address NIH
needs in scientific portfolio analysis
Office of Portfolio Analysis (OPA) DPCPSI OD 12
Outcomes of the Portfolio Analysis Symposium
Links to the Symposium agenda and archived videocast are
available on the OPA web site
httpdpcpsinihgovopaindexaspx
Outside experts presented new tools and approaches in each
topic one of these tools (Sci2) was particularly well received
by Symposium attendees and is being added to QVR
The Symposium resulted in several collaborations Examples
include OPA initiatives in building tools for co-author network
analysis analysis of patent development and tracking of
patent licenses
Office of Portfolio Analysis (OPA) DPCPSI OD 13
The Office of Portfolio Analysis (OPA)
OPA was established in 2011 to provide multiple services
Scientific portfolio analysishellip
o of all Common Fund initiatives
o in response to requests by senior leadership
Coordinate portfolio analysis activities across NIH
Train NIH staff to promote the effective use of analytical tools
o Regularly scheduled courses
o Ad hoc consultations
Improve portfolio analysis at NIH
Office of Portfolio Analysis (OPA) DPCPSI OD 14
Build Computational Tools and Train NIH Staff
Office of Portfolio Analysis (OPA) DPCPSI OD
OPA
TOOLS
LAB
12 laptops
Two workstations Building 1 Room B301
Smartboard
Linux server
Videoconferencing capabilities
15
OPA Consultations with NIH Staff
IC Contact Activity
NCCAM John Williamson Consult
NCI John Hewes Consult amp Collaboration
NCI Tanya Agurs-Collins Consult
NHLBI Marc Charette Consult amp Collaboration
NIAMS Faye Chen Consult
NIAMS Anita Linde Consult
NIDDK Lisa Spain Consult
NIGMS Michelle Hamlet Consult
NIGMS Ward Smith Consult
NIGMS Kelley Smith Consult amp Training
NIMH Yancy Bodenstein Consult amp Collaboration
NIMHD Ligia Artiles Consult amp Training
NLM Alan Vanbierlet Consult ampTraining
Office of Portfolio Analysis (OPA) DPCPSI OD 16
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 17
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Scientific Portfolio Analysis at NIH
Data-driven approaches to program development
What is the best way to accelerate scientific progress
An example
OPA analysis of 38 Metabolomics Centers in the US
most of which receive funding from NIH (~$70M in 2010)
Office of Portfolio Analysis (OPA) DPCPSI OD 4
$225M Global Investment in Metabolomics in FY10
bull Netherlands Metabolomics Center $67M
bull BBSRC UK Plant and Microbial Metabolomics $104M
bull Canadian Human Metabolome Database $81M
Office of Portfolio Analysis (OPA) DPCPSI OD 5
Europe
Japan
FY09 Metabolomics Co-authorship Networks
USA
Office of Portfolio Analysis (OPA) DPCPSI OD 6
Scientific Portfolio Analysis at NIH
What is the best way to accelerate scientific progress
OPA analysis of 38 Metabolomics Centers in the US most of which receive funding from NIH (~$70M in 2010)
These Centers have overlapping goals yet operated in isolation
with little coordination or collaboration
Portfolio analysis shaped a new Metabolomics RFA aimed at
improving coordination and leveraging existing resources
Office of Portfolio Analysis (OPA) DPCPSI OD 7
o httpdpcpsinihgovportfolio_analysis
Coordination of Portfolio Analysis Efforts at NIH
Portfolio Analysis Workshop (February 6 Natcher Conf Ctr)
Portfolio Analysis Symposium (July 23-24 Natcher Conf Ctr)
Build a computer lab to tailor existing and new computational tools to
NIH needs and to train NIH staff in their use
Centralized web-based repository to disseminate computational tools
Standing trans-NIH Working Group
Office of Portfolio Analysis (OPA) DPCPSI OD 8
Goals of the Portfolio Analysis Workshop Feb 5-6 2012
Discuss perceived needs in portfolio analysis
Use needs assessment to plan the Portfolio Analysis Symposium
(July 23-24 Natcher Conf Ctr)
Cover topics of broad interest to NIH decision-makers including
o Strategic planning
o Uses of portfolio analysis
o Overlap in NIH portfolios
o Measuring impact
o New portfolio analysis tools
o Identification of emerging areas
Office of Portfolio Analysis (OPA) DPCPSI OD 9
90
80
70
60
50
40
30
20
10
0
Outcomes of the Portfolio Analysis Workshop
o Full registration within 48 hrs of the announcement
o gt500 participants ndash approved for ESA training credit
Office of Portfolio Analysis (OPA) DPCPSI OD 10
Outcomes of the Portfolio Analysis Workshop
Survey results
Highest priority Topics included
o Measuring impact 47
o Gaps and overlap in NIH portfolios 30
o Identification of emerging areas 18
o Categorizing portfolios 5
What needs should OPA try to address
o Build better tools easier to use tools 65
o Provide training and support 50
o Develop targeted case studies 15
Office of Portfolio Analysis (OPA) DPCPSI OD 11
Goals of the Portfolio Analysis Symposium July 23-24 2012
Bring outside experts from academia government and the
private sector to discuss and demo state-of-the-art
approaches in scientific portfolio analysis
Choose those with expertise in areas identified as critical and
of greatest interest to portfolio analysis stakeholders at NIH
o Measuring Impact
o Identifying Gaps and Impact
o Identifying Emerging Areas
Facilitate the development of collaborations that address NIH
needs in scientific portfolio analysis
Office of Portfolio Analysis (OPA) DPCPSI OD 12
Outcomes of the Portfolio Analysis Symposium
Links to the Symposium agenda and archived videocast are
available on the OPA web site
httpdpcpsinihgovopaindexaspx
Outside experts presented new tools and approaches in each
topic one of these tools (Sci2) was particularly well received
by Symposium attendees and is being added to QVR
The Symposium resulted in several collaborations Examples
include OPA initiatives in building tools for co-author network
analysis analysis of patent development and tracking of
patent licenses
Office of Portfolio Analysis (OPA) DPCPSI OD 13
The Office of Portfolio Analysis (OPA)
OPA was established in 2011 to provide multiple services
Scientific portfolio analysishellip
o of all Common Fund initiatives
o in response to requests by senior leadership
Coordinate portfolio analysis activities across NIH
Train NIH staff to promote the effective use of analytical tools
o Regularly scheduled courses
o Ad hoc consultations
Improve portfolio analysis at NIH
Office of Portfolio Analysis (OPA) DPCPSI OD 14
Build Computational Tools and Train NIH Staff
Office of Portfolio Analysis (OPA) DPCPSI OD
OPA
TOOLS
LAB
12 laptops
Two workstations Building 1 Room B301
Smartboard
Linux server
Videoconferencing capabilities
15
OPA Consultations with NIH Staff
IC Contact Activity
NCCAM John Williamson Consult
NCI John Hewes Consult amp Collaboration
NCI Tanya Agurs-Collins Consult
NHLBI Marc Charette Consult amp Collaboration
NIAMS Faye Chen Consult
NIAMS Anita Linde Consult
NIDDK Lisa Spain Consult
NIGMS Michelle Hamlet Consult
NIGMS Ward Smith Consult
NIGMS Kelley Smith Consult amp Training
NIMH Yancy Bodenstein Consult amp Collaboration
NIMHD Ligia Artiles Consult amp Training
NLM Alan Vanbierlet Consult ampTraining
Office of Portfolio Analysis (OPA) DPCPSI OD 16
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 17
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
$225M Global Investment in Metabolomics in FY10
bull Netherlands Metabolomics Center $67M
bull BBSRC UK Plant and Microbial Metabolomics $104M
bull Canadian Human Metabolome Database $81M
Office of Portfolio Analysis (OPA) DPCPSI OD 5
Europe
Japan
FY09 Metabolomics Co-authorship Networks
USA
Office of Portfolio Analysis (OPA) DPCPSI OD 6
Scientific Portfolio Analysis at NIH
What is the best way to accelerate scientific progress
OPA analysis of 38 Metabolomics Centers in the US most of which receive funding from NIH (~$70M in 2010)
These Centers have overlapping goals yet operated in isolation
with little coordination or collaboration
Portfolio analysis shaped a new Metabolomics RFA aimed at
improving coordination and leveraging existing resources
Office of Portfolio Analysis (OPA) DPCPSI OD 7
o httpdpcpsinihgovportfolio_analysis
Coordination of Portfolio Analysis Efforts at NIH
Portfolio Analysis Workshop (February 6 Natcher Conf Ctr)
Portfolio Analysis Symposium (July 23-24 Natcher Conf Ctr)
Build a computer lab to tailor existing and new computational tools to
NIH needs and to train NIH staff in their use
Centralized web-based repository to disseminate computational tools
Standing trans-NIH Working Group
Office of Portfolio Analysis (OPA) DPCPSI OD 8
Goals of the Portfolio Analysis Workshop Feb 5-6 2012
Discuss perceived needs in portfolio analysis
Use needs assessment to plan the Portfolio Analysis Symposium
(July 23-24 Natcher Conf Ctr)
Cover topics of broad interest to NIH decision-makers including
o Strategic planning
o Uses of portfolio analysis
o Overlap in NIH portfolios
o Measuring impact
o New portfolio analysis tools
o Identification of emerging areas
Office of Portfolio Analysis (OPA) DPCPSI OD 9
90
80
70
60
50
40
30
20
10
0
Outcomes of the Portfolio Analysis Workshop
o Full registration within 48 hrs of the announcement
o gt500 participants ndash approved for ESA training credit
Office of Portfolio Analysis (OPA) DPCPSI OD 10
Outcomes of the Portfolio Analysis Workshop
Survey results
Highest priority Topics included
o Measuring impact 47
o Gaps and overlap in NIH portfolios 30
o Identification of emerging areas 18
o Categorizing portfolios 5
What needs should OPA try to address
o Build better tools easier to use tools 65
o Provide training and support 50
o Develop targeted case studies 15
Office of Portfolio Analysis (OPA) DPCPSI OD 11
Goals of the Portfolio Analysis Symposium July 23-24 2012
Bring outside experts from academia government and the
private sector to discuss and demo state-of-the-art
approaches in scientific portfolio analysis
Choose those with expertise in areas identified as critical and
of greatest interest to portfolio analysis stakeholders at NIH
o Measuring Impact
o Identifying Gaps and Impact
o Identifying Emerging Areas
Facilitate the development of collaborations that address NIH
needs in scientific portfolio analysis
Office of Portfolio Analysis (OPA) DPCPSI OD 12
Outcomes of the Portfolio Analysis Symposium
Links to the Symposium agenda and archived videocast are
available on the OPA web site
httpdpcpsinihgovopaindexaspx
Outside experts presented new tools and approaches in each
topic one of these tools (Sci2) was particularly well received
by Symposium attendees and is being added to QVR
The Symposium resulted in several collaborations Examples
include OPA initiatives in building tools for co-author network
analysis analysis of patent development and tracking of
patent licenses
Office of Portfolio Analysis (OPA) DPCPSI OD 13
The Office of Portfolio Analysis (OPA)
OPA was established in 2011 to provide multiple services
Scientific portfolio analysishellip
o of all Common Fund initiatives
o in response to requests by senior leadership
Coordinate portfolio analysis activities across NIH
Train NIH staff to promote the effective use of analytical tools
o Regularly scheduled courses
o Ad hoc consultations
Improve portfolio analysis at NIH
Office of Portfolio Analysis (OPA) DPCPSI OD 14
Build Computational Tools and Train NIH Staff
Office of Portfolio Analysis (OPA) DPCPSI OD
OPA
TOOLS
LAB
12 laptops
Two workstations Building 1 Room B301
Smartboard
Linux server
Videoconferencing capabilities
15
OPA Consultations with NIH Staff
IC Contact Activity
NCCAM John Williamson Consult
NCI John Hewes Consult amp Collaboration
NCI Tanya Agurs-Collins Consult
NHLBI Marc Charette Consult amp Collaboration
NIAMS Faye Chen Consult
NIAMS Anita Linde Consult
NIDDK Lisa Spain Consult
NIGMS Michelle Hamlet Consult
NIGMS Ward Smith Consult
NIGMS Kelley Smith Consult amp Training
NIMH Yancy Bodenstein Consult amp Collaboration
NIMHD Ligia Artiles Consult amp Training
NLM Alan Vanbierlet Consult ampTraining
Office of Portfolio Analysis (OPA) DPCPSI OD 16
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 17
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Europe
Japan
FY09 Metabolomics Co-authorship Networks
USA
Office of Portfolio Analysis (OPA) DPCPSI OD 6
Scientific Portfolio Analysis at NIH
What is the best way to accelerate scientific progress
OPA analysis of 38 Metabolomics Centers in the US most of which receive funding from NIH (~$70M in 2010)
These Centers have overlapping goals yet operated in isolation
with little coordination or collaboration
Portfolio analysis shaped a new Metabolomics RFA aimed at
improving coordination and leveraging existing resources
Office of Portfolio Analysis (OPA) DPCPSI OD 7
o httpdpcpsinihgovportfolio_analysis
Coordination of Portfolio Analysis Efforts at NIH
Portfolio Analysis Workshop (February 6 Natcher Conf Ctr)
Portfolio Analysis Symposium (July 23-24 Natcher Conf Ctr)
Build a computer lab to tailor existing and new computational tools to
NIH needs and to train NIH staff in their use
Centralized web-based repository to disseminate computational tools
Standing trans-NIH Working Group
Office of Portfolio Analysis (OPA) DPCPSI OD 8
Goals of the Portfolio Analysis Workshop Feb 5-6 2012
Discuss perceived needs in portfolio analysis
Use needs assessment to plan the Portfolio Analysis Symposium
(July 23-24 Natcher Conf Ctr)
Cover topics of broad interest to NIH decision-makers including
o Strategic planning
o Uses of portfolio analysis
o Overlap in NIH portfolios
o Measuring impact
o New portfolio analysis tools
o Identification of emerging areas
Office of Portfolio Analysis (OPA) DPCPSI OD 9
90
80
70
60
50
40
30
20
10
0
Outcomes of the Portfolio Analysis Workshop
o Full registration within 48 hrs of the announcement
o gt500 participants ndash approved for ESA training credit
Office of Portfolio Analysis (OPA) DPCPSI OD 10
Outcomes of the Portfolio Analysis Workshop
Survey results
Highest priority Topics included
o Measuring impact 47
o Gaps and overlap in NIH portfolios 30
o Identification of emerging areas 18
o Categorizing portfolios 5
What needs should OPA try to address
o Build better tools easier to use tools 65
o Provide training and support 50
o Develop targeted case studies 15
Office of Portfolio Analysis (OPA) DPCPSI OD 11
Goals of the Portfolio Analysis Symposium July 23-24 2012
Bring outside experts from academia government and the
private sector to discuss and demo state-of-the-art
approaches in scientific portfolio analysis
Choose those with expertise in areas identified as critical and
of greatest interest to portfolio analysis stakeholders at NIH
o Measuring Impact
o Identifying Gaps and Impact
o Identifying Emerging Areas
Facilitate the development of collaborations that address NIH
needs in scientific portfolio analysis
Office of Portfolio Analysis (OPA) DPCPSI OD 12
Outcomes of the Portfolio Analysis Symposium
Links to the Symposium agenda and archived videocast are
available on the OPA web site
httpdpcpsinihgovopaindexaspx
Outside experts presented new tools and approaches in each
topic one of these tools (Sci2) was particularly well received
by Symposium attendees and is being added to QVR
The Symposium resulted in several collaborations Examples
include OPA initiatives in building tools for co-author network
analysis analysis of patent development and tracking of
patent licenses
Office of Portfolio Analysis (OPA) DPCPSI OD 13
The Office of Portfolio Analysis (OPA)
OPA was established in 2011 to provide multiple services
Scientific portfolio analysishellip
o of all Common Fund initiatives
o in response to requests by senior leadership
Coordinate portfolio analysis activities across NIH
Train NIH staff to promote the effective use of analytical tools
o Regularly scheduled courses
o Ad hoc consultations
Improve portfolio analysis at NIH
Office of Portfolio Analysis (OPA) DPCPSI OD 14
Build Computational Tools and Train NIH Staff
Office of Portfolio Analysis (OPA) DPCPSI OD
OPA
TOOLS
LAB
12 laptops
Two workstations Building 1 Room B301
Smartboard
Linux server
Videoconferencing capabilities
15
OPA Consultations with NIH Staff
IC Contact Activity
NCCAM John Williamson Consult
NCI John Hewes Consult amp Collaboration
NCI Tanya Agurs-Collins Consult
NHLBI Marc Charette Consult amp Collaboration
NIAMS Faye Chen Consult
NIAMS Anita Linde Consult
NIDDK Lisa Spain Consult
NIGMS Michelle Hamlet Consult
NIGMS Ward Smith Consult
NIGMS Kelley Smith Consult amp Training
NIMH Yancy Bodenstein Consult amp Collaboration
NIMHD Ligia Artiles Consult amp Training
NLM Alan Vanbierlet Consult ampTraining
Office of Portfolio Analysis (OPA) DPCPSI OD 16
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 17
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Scientific Portfolio Analysis at NIH
What is the best way to accelerate scientific progress
OPA analysis of 38 Metabolomics Centers in the US most of which receive funding from NIH (~$70M in 2010)
These Centers have overlapping goals yet operated in isolation
with little coordination or collaboration
Portfolio analysis shaped a new Metabolomics RFA aimed at
improving coordination and leveraging existing resources
Office of Portfolio Analysis (OPA) DPCPSI OD 7
o httpdpcpsinihgovportfolio_analysis
Coordination of Portfolio Analysis Efforts at NIH
Portfolio Analysis Workshop (February 6 Natcher Conf Ctr)
Portfolio Analysis Symposium (July 23-24 Natcher Conf Ctr)
Build a computer lab to tailor existing and new computational tools to
NIH needs and to train NIH staff in their use
Centralized web-based repository to disseminate computational tools
Standing trans-NIH Working Group
Office of Portfolio Analysis (OPA) DPCPSI OD 8
Goals of the Portfolio Analysis Workshop Feb 5-6 2012
Discuss perceived needs in portfolio analysis
Use needs assessment to plan the Portfolio Analysis Symposium
(July 23-24 Natcher Conf Ctr)
Cover topics of broad interest to NIH decision-makers including
o Strategic planning
o Uses of portfolio analysis
o Overlap in NIH portfolios
o Measuring impact
o New portfolio analysis tools
o Identification of emerging areas
Office of Portfolio Analysis (OPA) DPCPSI OD 9
90
80
70
60
50
40
30
20
10
0
Outcomes of the Portfolio Analysis Workshop
o Full registration within 48 hrs of the announcement
o gt500 participants ndash approved for ESA training credit
Office of Portfolio Analysis (OPA) DPCPSI OD 10
Outcomes of the Portfolio Analysis Workshop
Survey results
Highest priority Topics included
o Measuring impact 47
o Gaps and overlap in NIH portfolios 30
o Identification of emerging areas 18
o Categorizing portfolios 5
What needs should OPA try to address
o Build better tools easier to use tools 65
o Provide training and support 50
o Develop targeted case studies 15
Office of Portfolio Analysis (OPA) DPCPSI OD 11
Goals of the Portfolio Analysis Symposium July 23-24 2012
Bring outside experts from academia government and the
private sector to discuss and demo state-of-the-art
approaches in scientific portfolio analysis
Choose those with expertise in areas identified as critical and
of greatest interest to portfolio analysis stakeholders at NIH
o Measuring Impact
o Identifying Gaps and Impact
o Identifying Emerging Areas
Facilitate the development of collaborations that address NIH
needs in scientific portfolio analysis
Office of Portfolio Analysis (OPA) DPCPSI OD 12
Outcomes of the Portfolio Analysis Symposium
Links to the Symposium agenda and archived videocast are
available on the OPA web site
httpdpcpsinihgovopaindexaspx
Outside experts presented new tools and approaches in each
topic one of these tools (Sci2) was particularly well received
by Symposium attendees and is being added to QVR
The Symposium resulted in several collaborations Examples
include OPA initiatives in building tools for co-author network
analysis analysis of patent development and tracking of
patent licenses
Office of Portfolio Analysis (OPA) DPCPSI OD 13
The Office of Portfolio Analysis (OPA)
OPA was established in 2011 to provide multiple services
Scientific portfolio analysishellip
o of all Common Fund initiatives
o in response to requests by senior leadership
Coordinate portfolio analysis activities across NIH
Train NIH staff to promote the effective use of analytical tools
o Regularly scheduled courses
o Ad hoc consultations
Improve portfolio analysis at NIH
Office of Portfolio Analysis (OPA) DPCPSI OD 14
Build Computational Tools and Train NIH Staff
Office of Portfolio Analysis (OPA) DPCPSI OD
OPA
TOOLS
LAB
12 laptops
Two workstations Building 1 Room B301
Smartboard
Linux server
Videoconferencing capabilities
15
OPA Consultations with NIH Staff
IC Contact Activity
NCCAM John Williamson Consult
NCI John Hewes Consult amp Collaboration
NCI Tanya Agurs-Collins Consult
NHLBI Marc Charette Consult amp Collaboration
NIAMS Faye Chen Consult
NIAMS Anita Linde Consult
NIDDK Lisa Spain Consult
NIGMS Michelle Hamlet Consult
NIGMS Ward Smith Consult
NIGMS Kelley Smith Consult amp Training
NIMH Yancy Bodenstein Consult amp Collaboration
NIMHD Ligia Artiles Consult amp Training
NLM Alan Vanbierlet Consult ampTraining
Office of Portfolio Analysis (OPA) DPCPSI OD 16
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 17
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
o httpdpcpsinihgovportfolio_analysis
Coordination of Portfolio Analysis Efforts at NIH
Portfolio Analysis Workshop (February 6 Natcher Conf Ctr)
Portfolio Analysis Symposium (July 23-24 Natcher Conf Ctr)
Build a computer lab to tailor existing and new computational tools to
NIH needs and to train NIH staff in their use
Centralized web-based repository to disseminate computational tools
Standing trans-NIH Working Group
Office of Portfolio Analysis (OPA) DPCPSI OD 8
Goals of the Portfolio Analysis Workshop Feb 5-6 2012
Discuss perceived needs in portfolio analysis
Use needs assessment to plan the Portfolio Analysis Symposium
(July 23-24 Natcher Conf Ctr)
Cover topics of broad interest to NIH decision-makers including
o Strategic planning
o Uses of portfolio analysis
o Overlap in NIH portfolios
o Measuring impact
o New portfolio analysis tools
o Identification of emerging areas
Office of Portfolio Analysis (OPA) DPCPSI OD 9
90
80
70
60
50
40
30
20
10
0
Outcomes of the Portfolio Analysis Workshop
o Full registration within 48 hrs of the announcement
o gt500 participants ndash approved for ESA training credit
Office of Portfolio Analysis (OPA) DPCPSI OD 10
Outcomes of the Portfolio Analysis Workshop
Survey results
Highest priority Topics included
o Measuring impact 47
o Gaps and overlap in NIH portfolios 30
o Identification of emerging areas 18
o Categorizing portfolios 5
What needs should OPA try to address
o Build better tools easier to use tools 65
o Provide training and support 50
o Develop targeted case studies 15
Office of Portfolio Analysis (OPA) DPCPSI OD 11
Goals of the Portfolio Analysis Symposium July 23-24 2012
Bring outside experts from academia government and the
private sector to discuss and demo state-of-the-art
approaches in scientific portfolio analysis
Choose those with expertise in areas identified as critical and
of greatest interest to portfolio analysis stakeholders at NIH
o Measuring Impact
o Identifying Gaps and Impact
o Identifying Emerging Areas
Facilitate the development of collaborations that address NIH
needs in scientific portfolio analysis
Office of Portfolio Analysis (OPA) DPCPSI OD 12
Outcomes of the Portfolio Analysis Symposium
Links to the Symposium agenda and archived videocast are
available on the OPA web site
httpdpcpsinihgovopaindexaspx
Outside experts presented new tools and approaches in each
topic one of these tools (Sci2) was particularly well received
by Symposium attendees and is being added to QVR
The Symposium resulted in several collaborations Examples
include OPA initiatives in building tools for co-author network
analysis analysis of patent development and tracking of
patent licenses
Office of Portfolio Analysis (OPA) DPCPSI OD 13
The Office of Portfolio Analysis (OPA)
OPA was established in 2011 to provide multiple services
Scientific portfolio analysishellip
o of all Common Fund initiatives
o in response to requests by senior leadership
Coordinate portfolio analysis activities across NIH
Train NIH staff to promote the effective use of analytical tools
o Regularly scheduled courses
o Ad hoc consultations
Improve portfolio analysis at NIH
Office of Portfolio Analysis (OPA) DPCPSI OD 14
Build Computational Tools and Train NIH Staff
Office of Portfolio Analysis (OPA) DPCPSI OD
OPA
TOOLS
LAB
12 laptops
Two workstations Building 1 Room B301
Smartboard
Linux server
Videoconferencing capabilities
15
OPA Consultations with NIH Staff
IC Contact Activity
NCCAM John Williamson Consult
NCI John Hewes Consult amp Collaboration
NCI Tanya Agurs-Collins Consult
NHLBI Marc Charette Consult amp Collaboration
NIAMS Faye Chen Consult
NIAMS Anita Linde Consult
NIDDK Lisa Spain Consult
NIGMS Michelle Hamlet Consult
NIGMS Ward Smith Consult
NIGMS Kelley Smith Consult amp Training
NIMH Yancy Bodenstein Consult amp Collaboration
NIMHD Ligia Artiles Consult amp Training
NLM Alan Vanbierlet Consult ampTraining
Office of Portfolio Analysis (OPA) DPCPSI OD 16
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 17
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Goals of the Portfolio Analysis Workshop Feb 5-6 2012
Discuss perceived needs in portfolio analysis
Use needs assessment to plan the Portfolio Analysis Symposium
(July 23-24 Natcher Conf Ctr)
Cover topics of broad interest to NIH decision-makers including
o Strategic planning
o Uses of portfolio analysis
o Overlap in NIH portfolios
o Measuring impact
o New portfolio analysis tools
o Identification of emerging areas
Office of Portfolio Analysis (OPA) DPCPSI OD 9
90
80
70
60
50
40
30
20
10
0
Outcomes of the Portfolio Analysis Workshop
o Full registration within 48 hrs of the announcement
o gt500 participants ndash approved for ESA training credit
Office of Portfolio Analysis (OPA) DPCPSI OD 10
Outcomes of the Portfolio Analysis Workshop
Survey results
Highest priority Topics included
o Measuring impact 47
o Gaps and overlap in NIH portfolios 30
o Identification of emerging areas 18
o Categorizing portfolios 5
What needs should OPA try to address
o Build better tools easier to use tools 65
o Provide training and support 50
o Develop targeted case studies 15
Office of Portfolio Analysis (OPA) DPCPSI OD 11
Goals of the Portfolio Analysis Symposium July 23-24 2012
Bring outside experts from academia government and the
private sector to discuss and demo state-of-the-art
approaches in scientific portfolio analysis
Choose those with expertise in areas identified as critical and
of greatest interest to portfolio analysis stakeholders at NIH
o Measuring Impact
o Identifying Gaps and Impact
o Identifying Emerging Areas
Facilitate the development of collaborations that address NIH
needs in scientific portfolio analysis
Office of Portfolio Analysis (OPA) DPCPSI OD 12
Outcomes of the Portfolio Analysis Symposium
Links to the Symposium agenda and archived videocast are
available on the OPA web site
httpdpcpsinihgovopaindexaspx
Outside experts presented new tools and approaches in each
topic one of these tools (Sci2) was particularly well received
by Symposium attendees and is being added to QVR
The Symposium resulted in several collaborations Examples
include OPA initiatives in building tools for co-author network
analysis analysis of patent development and tracking of
patent licenses
Office of Portfolio Analysis (OPA) DPCPSI OD 13
The Office of Portfolio Analysis (OPA)
OPA was established in 2011 to provide multiple services
Scientific portfolio analysishellip
o of all Common Fund initiatives
o in response to requests by senior leadership
Coordinate portfolio analysis activities across NIH
Train NIH staff to promote the effective use of analytical tools
o Regularly scheduled courses
o Ad hoc consultations
Improve portfolio analysis at NIH
Office of Portfolio Analysis (OPA) DPCPSI OD 14
Build Computational Tools and Train NIH Staff
Office of Portfolio Analysis (OPA) DPCPSI OD
OPA
TOOLS
LAB
12 laptops
Two workstations Building 1 Room B301
Smartboard
Linux server
Videoconferencing capabilities
15
OPA Consultations with NIH Staff
IC Contact Activity
NCCAM John Williamson Consult
NCI John Hewes Consult amp Collaboration
NCI Tanya Agurs-Collins Consult
NHLBI Marc Charette Consult amp Collaboration
NIAMS Faye Chen Consult
NIAMS Anita Linde Consult
NIDDK Lisa Spain Consult
NIGMS Michelle Hamlet Consult
NIGMS Ward Smith Consult
NIGMS Kelley Smith Consult amp Training
NIMH Yancy Bodenstein Consult amp Collaboration
NIMHD Ligia Artiles Consult amp Training
NLM Alan Vanbierlet Consult ampTraining
Office of Portfolio Analysis (OPA) DPCPSI OD 16
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 17
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
90
80
70
60
50
40
30
20
10
0
Outcomes of the Portfolio Analysis Workshop
o Full registration within 48 hrs of the announcement
o gt500 participants ndash approved for ESA training credit
Office of Portfolio Analysis (OPA) DPCPSI OD 10
Outcomes of the Portfolio Analysis Workshop
Survey results
Highest priority Topics included
o Measuring impact 47
o Gaps and overlap in NIH portfolios 30
o Identification of emerging areas 18
o Categorizing portfolios 5
What needs should OPA try to address
o Build better tools easier to use tools 65
o Provide training and support 50
o Develop targeted case studies 15
Office of Portfolio Analysis (OPA) DPCPSI OD 11
Goals of the Portfolio Analysis Symposium July 23-24 2012
Bring outside experts from academia government and the
private sector to discuss and demo state-of-the-art
approaches in scientific portfolio analysis
Choose those with expertise in areas identified as critical and
of greatest interest to portfolio analysis stakeholders at NIH
o Measuring Impact
o Identifying Gaps and Impact
o Identifying Emerging Areas
Facilitate the development of collaborations that address NIH
needs in scientific portfolio analysis
Office of Portfolio Analysis (OPA) DPCPSI OD 12
Outcomes of the Portfolio Analysis Symposium
Links to the Symposium agenda and archived videocast are
available on the OPA web site
httpdpcpsinihgovopaindexaspx
Outside experts presented new tools and approaches in each
topic one of these tools (Sci2) was particularly well received
by Symposium attendees and is being added to QVR
The Symposium resulted in several collaborations Examples
include OPA initiatives in building tools for co-author network
analysis analysis of patent development and tracking of
patent licenses
Office of Portfolio Analysis (OPA) DPCPSI OD 13
The Office of Portfolio Analysis (OPA)
OPA was established in 2011 to provide multiple services
Scientific portfolio analysishellip
o of all Common Fund initiatives
o in response to requests by senior leadership
Coordinate portfolio analysis activities across NIH
Train NIH staff to promote the effective use of analytical tools
o Regularly scheduled courses
o Ad hoc consultations
Improve portfolio analysis at NIH
Office of Portfolio Analysis (OPA) DPCPSI OD 14
Build Computational Tools and Train NIH Staff
Office of Portfolio Analysis (OPA) DPCPSI OD
OPA
TOOLS
LAB
12 laptops
Two workstations Building 1 Room B301
Smartboard
Linux server
Videoconferencing capabilities
15
OPA Consultations with NIH Staff
IC Contact Activity
NCCAM John Williamson Consult
NCI John Hewes Consult amp Collaboration
NCI Tanya Agurs-Collins Consult
NHLBI Marc Charette Consult amp Collaboration
NIAMS Faye Chen Consult
NIAMS Anita Linde Consult
NIDDK Lisa Spain Consult
NIGMS Michelle Hamlet Consult
NIGMS Ward Smith Consult
NIGMS Kelley Smith Consult amp Training
NIMH Yancy Bodenstein Consult amp Collaboration
NIMHD Ligia Artiles Consult amp Training
NLM Alan Vanbierlet Consult ampTraining
Office of Portfolio Analysis (OPA) DPCPSI OD 16
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 17
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Outcomes of the Portfolio Analysis Workshop
Survey results
Highest priority Topics included
o Measuring impact 47
o Gaps and overlap in NIH portfolios 30
o Identification of emerging areas 18
o Categorizing portfolios 5
What needs should OPA try to address
o Build better tools easier to use tools 65
o Provide training and support 50
o Develop targeted case studies 15
Office of Portfolio Analysis (OPA) DPCPSI OD 11
Goals of the Portfolio Analysis Symposium July 23-24 2012
Bring outside experts from academia government and the
private sector to discuss and demo state-of-the-art
approaches in scientific portfolio analysis
Choose those with expertise in areas identified as critical and
of greatest interest to portfolio analysis stakeholders at NIH
o Measuring Impact
o Identifying Gaps and Impact
o Identifying Emerging Areas
Facilitate the development of collaborations that address NIH
needs in scientific portfolio analysis
Office of Portfolio Analysis (OPA) DPCPSI OD 12
Outcomes of the Portfolio Analysis Symposium
Links to the Symposium agenda and archived videocast are
available on the OPA web site
httpdpcpsinihgovopaindexaspx
Outside experts presented new tools and approaches in each
topic one of these tools (Sci2) was particularly well received
by Symposium attendees and is being added to QVR
The Symposium resulted in several collaborations Examples
include OPA initiatives in building tools for co-author network
analysis analysis of patent development and tracking of
patent licenses
Office of Portfolio Analysis (OPA) DPCPSI OD 13
The Office of Portfolio Analysis (OPA)
OPA was established in 2011 to provide multiple services
Scientific portfolio analysishellip
o of all Common Fund initiatives
o in response to requests by senior leadership
Coordinate portfolio analysis activities across NIH
Train NIH staff to promote the effective use of analytical tools
o Regularly scheduled courses
o Ad hoc consultations
Improve portfolio analysis at NIH
Office of Portfolio Analysis (OPA) DPCPSI OD 14
Build Computational Tools and Train NIH Staff
Office of Portfolio Analysis (OPA) DPCPSI OD
OPA
TOOLS
LAB
12 laptops
Two workstations Building 1 Room B301
Smartboard
Linux server
Videoconferencing capabilities
15
OPA Consultations with NIH Staff
IC Contact Activity
NCCAM John Williamson Consult
NCI John Hewes Consult amp Collaboration
NCI Tanya Agurs-Collins Consult
NHLBI Marc Charette Consult amp Collaboration
NIAMS Faye Chen Consult
NIAMS Anita Linde Consult
NIDDK Lisa Spain Consult
NIGMS Michelle Hamlet Consult
NIGMS Ward Smith Consult
NIGMS Kelley Smith Consult amp Training
NIMH Yancy Bodenstein Consult amp Collaboration
NIMHD Ligia Artiles Consult amp Training
NLM Alan Vanbierlet Consult ampTraining
Office of Portfolio Analysis (OPA) DPCPSI OD 16
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 17
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Goals of the Portfolio Analysis Symposium July 23-24 2012
Bring outside experts from academia government and the
private sector to discuss and demo state-of-the-art
approaches in scientific portfolio analysis
Choose those with expertise in areas identified as critical and
of greatest interest to portfolio analysis stakeholders at NIH
o Measuring Impact
o Identifying Gaps and Impact
o Identifying Emerging Areas
Facilitate the development of collaborations that address NIH
needs in scientific portfolio analysis
Office of Portfolio Analysis (OPA) DPCPSI OD 12
Outcomes of the Portfolio Analysis Symposium
Links to the Symposium agenda and archived videocast are
available on the OPA web site
httpdpcpsinihgovopaindexaspx
Outside experts presented new tools and approaches in each
topic one of these tools (Sci2) was particularly well received
by Symposium attendees and is being added to QVR
The Symposium resulted in several collaborations Examples
include OPA initiatives in building tools for co-author network
analysis analysis of patent development and tracking of
patent licenses
Office of Portfolio Analysis (OPA) DPCPSI OD 13
The Office of Portfolio Analysis (OPA)
OPA was established in 2011 to provide multiple services
Scientific portfolio analysishellip
o of all Common Fund initiatives
o in response to requests by senior leadership
Coordinate portfolio analysis activities across NIH
Train NIH staff to promote the effective use of analytical tools
o Regularly scheduled courses
o Ad hoc consultations
Improve portfolio analysis at NIH
Office of Portfolio Analysis (OPA) DPCPSI OD 14
Build Computational Tools and Train NIH Staff
Office of Portfolio Analysis (OPA) DPCPSI OD
OPA
TOOLS
LAB
12 laptops
Two workstations Building 1 Room B301
Smartboard
Linux server
Videoconferencing capabilities
15
OPA Consultations with NIH Staff
IC Contact Activity
NCCAM John Williamson Consult
NCI John Hewes Consult amp Collaboration
NCI Tanya Agurs-Collins Consult
NHLBI Marc Charette Consult amp Collaboration
NIAMS Faye Chen Consult
NIAMS Anita Linde Consult
NIDDK Lisa Spain Consult
NIGMS Michelle Hamlet Consult
NIGMS Ward Smith Consult
NIGMS Kelley Smith Consult amp Training
NIMH Yancy Bodenstein Consult amp Collaboration
NIMHD Ligia Artiles Consult amp Training
NLM Alan Vanbierlet Consult ampTraining
Office of Portfolio Analysis (OPA) DPCPSI OD 16
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 17
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Outcomes of the Portfolio Analysis Symposium
Links to the Symposium agenda and archived videocast are
available on the OPA web site
httpdpcpsinihgovopaindexaspx
Outside experts presented new tools and approaches in each
topic one of these tools (Sci2) was particularly well received
by Symposium attendees and is being added to QVR
The Symposium resulted in several collaborations Examples
include OPA initiatives in building tools for co-author network
analysis analysis of patent development and tracking of
patent licenses
Office of Portfolio Analysis (OPA) DPCPSI OD 13
The Office of Portfolio Analysis (OPA)
OPA was established in 2011 to provide multiple services
Scientific portfolio analysishellip
o of all Common Fund initiatives
o in response to requests by senior leadership
Coordinate portfolio analysis activities across NIH
Train NIH staff to promote the effective use of analytical tools
o Regularly scheduled courses
o Ad hoc consultations
Improve portfolio analysis at NIH
Office of Portfolio Analysis (OPA) DPCPSI OD 14
Build Computational Tools and Train NIH Staff
Office of Portfolio Analysis (OPA) DPCPSI OD
OPA
TOOLS
LAB
12 laptops
Two workstations Building 1 Room B301
Smartboard
Linux server
Videoconferencing capabilities
15
OPA Consultations with NIH Staff
IC Contact Activity
NCCAM John Williamson Consult
NCI John Hewes Consult amp Collaboration
NCI Tanya Agurs-Collins Consult
NHLBI Marc Charette Consult amp Collaboration
NIAMS Faye Chen Consult
NIAMS Anita Linde Consult
NIDDK Lisa Spain Consult
NIGMS Michelle Hamlet Consult
NIGMS Ward Smith Consult
NIGMS Kelley Smith Consult amp Training
NIMH Yancy Bodenstein Consult amp Collaboration
NIMHD Ligia Artiles Consult amp Training
NLM Alan Vanbierlet Consult ampTraining
Office of Portfolio Analysis (OPA) DPCPSI OD 16
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 17
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
The Office of Portfolio Analysis (OPA)
OPA was established in 2011 to provide multiple services
Scientific portfolio analysishellip
o of all Common Fund initiatives
o in response to requests by senior leadership
Coordinate portfolio analysis activities across NIH
Train NIH staff to promote the effective use of analytical tools
o Regularly scheduled courses
o Ad hoc consultations
Improve portfolio analysis at NIH
Office of Portfolio Analysis (OPA) DPCPSI OD 14
Build Computational Tools and Train NIH Staff
Office of Portfolio Analysis (OPA) DPCPSI OD
OPA
TOOLS
LAB
12 laptops
Two workstations Building 1 Room B301
Smartboard
Linux server
Videoconferencing capabilities
15
OPA Consultations with NIH Staff
IC Contact Activity
NCCAM John Williamson Consult
NCI John Hewes Consult amp Collaboration
NCI Tanya Agurs-Collins Consult
NHLBI Marc Charette Consult amp Collaboration
NIAMS Faye Chen Consult
NIAMS Anita Linde Consult
NIDDK Lisa Spain Consult
NIGMS Michelle Hamlet Consult
NIGMS Ward Smith Consult
NIGMS Kelley Smith Consult amp Training
NIMH Yancy Bodenstein Consult amp Collaboration
NIMHD Ligia Artiles Consult amp Training
NLM Alan Vanbierlet Consult ampTraining
Office of Portfolio Analysis (OPA) DPCPSI OD 16
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 17
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Build Computational Tools and Train NIH Staff
Office of Portfolio Analysis (OPA) DPCPSI OD
OPA
TOOLS
LAB
12 laptops
Two workstations Building 1 Room B301
Smartboard
Linux server
Videoconferencing capabilities
15
OPA Consultations with NIH Staff
IC Contact Activity
NCCAM John Williamson Consult
NCI John Hewes Consult amp Collaboration
NCI Tanya Agurs-Collins Consult
NHLBI Marc Charette Consult amp Collaboration
NIAMS Faye Chen Consult
NIAMS Anita Linde Consult
NIDDK Lisa Spain Consult
NIGMS Michelle Hamlet Consult
NIGMS Ward Smith Consult
NIGMS Kelley Smith Consult amp Training
NIMH Yancy Bodenstein Consult amp Collaboration
NIMHD Ligia Artiles Consult amp Training
NLM Alan Vanbierlet Consult ampTraining
Office of Portfolio Analysis (OPA) DPCPSI OD 16
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 17
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
OPA Consultations with NIH Staff
IC Contact Activity
NCCAM John Williamson Consult
NCI John Hewes Consult amp Collaboration
NCI Tanya Agurs-Collins Consult
NHLBI Marc Charette Consult amp Collaboration
NIAMS Faye Chen Consult
NIAMS Anita Linde Consult
NIDDK Lisa Spain Consult
NIGMS Michelle Hamlet Consult
NIGMS Ward Smith Consult
NIGMS Kelley Smith Consult amp Training
NIMH Yancy Bodenstein Consult amp Collaboration
NIMHD Ligia Artiles Consult amp Training
NLM Alan Vanbierlet Consult ampTraining
Office of Portfolio Analysis (OPA) DPCPSI OD 16
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 17
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 17
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
DATABASE MANAGEMENT
Our current ability to track awards output and health impact is limited
Funding of clinical trials
o Inadequate linkage between ClinicalTrialsgov and IMPAC II data
o NCI and CIT are collaborating to address this problem
Patents and licensing of intellectual property
o No database exists for NIH-funded patents and university licensing
o OPA is addressing this in collaboration w NIH tech transfer officers
Output of awards
o Inadequate linkage between NIH awards and literaturecitation data
o OPA is developing a next-gen disambiguation tool
18Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
TOOL DEVELOPMENT
New tools are needed to provide NIH staff grant applicants et al
with a current and accurate picture of NIH investments
In collaboration with Calvin Johnsonrsquos group at CIT OPA has
developed several new tools
Examples include the Hierarchical Clustering amp Classifier tools
o Comparative Effectiveness Research (CER)
o NIGMS Technology Development
o Disruptive Proteomics Common Fund Analysis
o Similarity Matrix of Standing Study Sections (SRGs)
bull Characterize the assignment of applications to SRGs
bull Evaluate how SRGs are presented to PIs
Office of Portfolio Analysis (OPA) DPCPSI OD 19
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Cancer Genetics Study Section vs All 165 Study Sections 2011 3 cycles
Study Section Code Match Score
Cancer Genetics CG 1000
Cancer Molecular Pathobiology
Tumor Cell Biology
Molecular Oncogenesis
Tumor Progression and Metastasis
Basic Mechanisms of Cancer Therapeutics
Developmental Therapeutics
Genetics of Health and Disease
Tumor Microenvironment
Genomics Computational Biology and Technology
Cancer Biomarkers
Molecular Genetics B
ChemoDietary Prevention
Epidemiology of Cancer
Radiation Therapeutics and Biology
Drug Discovery and Molecular Pharmacology
Molecular Neurogenetics
Cancer Etiology
Molecular Genetics C
Molecular Genetics A
CAMP
TCB
MONC
TPM
BMCT
DT
GHD
TME
GCAT
CBSS
MGB
CDP
EPIC
RTB
DMP
MNG
CE
MGC
MGA
0509
0452
0445
0405
0397
0394
0394
0393
0365
0364
0362
0360
0353
0346
0339
0331
0326
0321
0288
+
Office of Portfolio Analysis (OPA) DPCPSI OD 20
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Neural Basis of Psychopathology Addictions and Sleep Disorders Study Section vs All 165 Study Sections
2011 3 cycles
Study Section Code Match Score
Neural Basis of Psychopathology Addictions and Sleep Disorders NPAS 1000
Adult Psychopathology and Disorders of Aging
Pathophysiological Basis of Mental Disorders and Addictions
Cognition and Perception
Biobehavioral Regulation Learning and Ethology
Cognitive Neuroscience
Clinical Neuroscience and Neurodegeneration
Child Psychopathology and Developmental Disabilities
Neurobiology of Motivated Behavior
Molecular Neuropharmacology and Signaling
Developmental Brain Disorders
Neurotoxicology and Alcohol
Genetics of Health and Disease
APDA
PMDA
CP
BRLE
COG
CNN
CPDD
NMB
MNPS
DBD
NAL
GHD
0528
0493
0398
0378
0368
0363
0353
0341
0339
0326
0294
0189
+
Office of Portfolio Analysis (OPA) DPCPSI OD 21
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Improve Portfolio Analysis at NIH
Database management
Modeling
NIH output Tool
development
Office of Portfolio Analysis (OPA) DPCPSI OD 22
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
MODELING NIH OUTPUT
INPUT $$
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 23
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
MODELING NIH OUTPUT Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
HEALTH amp ECONOMIC
IMPACT
Office of Portfolio Analysis (OPA) DPCPSI OD 24
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Citation Profile for 12 Selected Biomed Journals
Four journals from each 1200
category
High impact factor gt25
Medium IF 13-17
Low IF 1-6
All papers in those 12 journals
in a single year 2007 Tota
l of
citations
800
400
0 0 1000 2000 3000 4000 5000 6000
All citations of those papers
since 2007 Rank order of 2007 papers (least to most cited)
Office of Portfolio Analysis (OPA) DPCPSI OD 25
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Averages Apply Only to Gaussian Distributions P
erc
en
tag
e
Adapted from
MEJ Newman
Contemp Physics
46322-351 (2005)
Heights of males Speeds of cars
10000000107500E+065
Po
pu
lation
(m
illio
ns)
400E+064 Size of cities Size of cities P
op
ula
tion
1000000106
Power law100000105
300E+063 in California in California
f(x) = Cx-10000104
2200E+06 1000103
100102 (with a tail) 1100E+06
1010
0 1010
0 100 200 300 400 500 1 10 100 000E+00
Rank order (smallest to largest)
26Office of Portfolio Analysis (OPA) DPCPSI OD
1
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Power Law Distribution for a High IF Journal
HIGH IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 27
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Power Law Distribution for a Medium IF Journal
MEDIUM IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 28
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Power Law Distribution for a Low IF Journal
LOW IMPACT FACTOR JOURNAL 2007
Office of Portfolio Analysis (OPA) DPCPSI OD 29
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Modeling NIH output Advanced Bibliometrics
INPUT $$
R01
OUTPUT pubs patents etc
For NIH R01 awardees who published in those 12 journals in 2007
Most of the projects have been resubmitted (T2 applications)
What is the relationship if any between prior output and review outcomes
Office of Portfolio Analysis (OPA) DPCPSI OD 30
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Journal IF Affects R01 Fate More Than of Citations
Effect of Impact Factor
Effect of Citations
Rela
tive
pro
babili
ty o
f success (
)
+10
+
20
+
30
+
40
R01 T2 renewal applicants who published in selected 2007 journals (N = 1631)
Baseline +20 +40 +60 +80 +100
Journal Impact Factor Number of Citations
Office of Portfolio Analysis (OPA) DPCPSI OD 31
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Acknowledgments
THE OPA TEAM
Geetha Senthil PhD Science Analyst
Paula Fearon PhD Science Analyst
Jean Yuan PhD Science Analyst
Carole Christian PhD Health Science Administrator
Ehsan Haque MS Data Analyst
Chuck Lynch PhD IT Specialist
Kevin Small PhD Software Developer
COLLABORATORS
Calvin Johnson PhD Chief High Performance Computing amp Informatics CIT
Katy Borner PhD Victor H Yngve Professor of Information Science
Indiana University
Office of Portfolio Analysis (OPA) DPCPSI OD 32
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Cit
ati
on
s
The Citation Pattern is Established Soon After Publication
Papers from 1997 in the 12 journals analyzed HI25
20
15
10
5
0
Years after date of publication
MED
LOW
(average of four journals in each group)
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Office of Portfolio Analysis (OPA) DPCPSI OD 33
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Cell Science NEJM Nature Genes Dev
J Exp Med
J Clin Inv JBC Lancet Mol Pharm
Genetics FEBS Lett
The Citation Pattern is Established Soon After Publication
45 Papers from 1997 in the 12 journals analyzed
40 1998
199935
2000
200130 2002
2003
2004
25
20 2005
2006 15
2007
200810
2009
20105 2011
0 JOURNAL 1 2 3 4 1 2 3 4 1 2 3 4
Cit
atio
ns
IMPACT FACTOR HIGH MEDIUM LOW
Office of Portfolio Analysis (OPA) DPCPSI OD 34
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
0
50
100
150
200
Cit
ati
on
s
200
MEDIUM IF JOURNALS
150
100
50
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7
Years after date of publication
8 9 10 11 12 13 14
HIGH IF JOURNALS
The Citation Pattern is Established Soon After Publication
Office of Portfolio Analysis (OPA) DPCPSI OD 35
Office of Portfolio Analysis (OPA) DPCPSI OD
Office of Portfolio Analysis (OPA) DPCPSI OD