MARKUSH CLAIMS: REPRESENTATION,
SEARCH, ANALYSIS & CONSTRUCTION
Árpád Figyelmesi
27th ICIC International Conference for the Information Community
Nice 2015
Where Are We and How to Go Forward?
OVERVIEW
History and current state of Markush claims
Origins
Dr. Eugene A. Markush
1888 Budapest, Hungary
1968 New York, USA
US1506316A
”The process for manufacture of
dyes…”colorantshistory.org
Importance
• Between 1994 and 2013
• 3,704,996 US patent
• 468,262 with Markush claims
• Every eighth patent contains
Markush claims
Joseph J. Mallon, 2014
The real value in Patents not in Drugs…
You Need Good Markush Technology
or
Lot of manual work (with unavoidable mistakes)
Drug Discovery workflow
Find relevant
documents
Analyze prior art
and invent
something new
Create your
own Patents
Variation types
• Substituent variation
• Position variation
• Frequency variation
• Homology variation
• Variation inside variation (nested)
• Additional logical constraints
Patent Markush
• Nested R-groups
• Homologies
• Additional logical constraints
Combinatorial Library
• No nested R-groups
• No Homologies, Repeating
units and Position variation
Markush types
Markush chemical space size
Zhengwei Peng, 2014
Existing databases
Thomson Markush database (MMS)
2.4 million patents
1.6 million Markush structures
2 million specific compounds
CAS Markush database (MARPAT)
0.5 million patent
1 million Markush structures
REPRESENTATION
Markush representation techniques and challenges
● R-groups
● Atom lists
● Bond list
● Position variations
● Repeating units
● Homology groups
Markush Representation
US5948793A
Claimed structure represented
with multiple structures
Workarounds
CONSTRUCTION
Sketching and automatic generation of Markush structures
General structure editors
Markush Editor
R-group definitions
Tree view Scaffold
Structure checker
Nesting view & Preview
Markush Composer
Automatic Markush generation from compound list
DOCUMENT CURATION
Extracting Markush structures from Patents
Representing Covered Chemical Space
● Document processing (XML,
PDF, HTML)
● Name to structure
● NLP technologies
● OSR (CLiDE & OSRA)
Automatic Markush extraction
ChemProspektor
InfoChem
Theseus research project founded
by the Federal Ministry of
Economics and Technology
Dr. Josef Eiblmaier, ACS National Meeting, Philadelphia, August 19 - 22, 2012
ChemCurator Markush extraction view
Markush editor
Example structures
Annotated document
Selected structures
Structure checker
MARKUSH SEARCH & ANALYSIS
Understanding covered chemical space and comparison
Markush Search & Hit Visualization
● Substructure
● Full structure
● (Similarity)
● Hit visualization
● Non-Hit visualization
Markush Enumeration
• Full enumeration
• Random enumeration
• Partial enumeration
• Library size calculation
• Biased enumeration
• (Property distribution
characterization)
Markush Overlap
Overlapping chemical space
calculation
Results:
● Percentage of overlap
● Overlapping Markush
Benefits:
● No enumeration
● No size limitations
SUMMARY
Summary
• Important breakthroughs in Markush technology– Search & hit visualization
– Comparison
– Construction
– Curation
• Active development in challenging areas– Similarity search
– Characterization
THANK YOUÁrpád Figyelmesi
27th ICIC International Conference for the Information Community
Nice 2015
Top Related