The iPlant Collaborative Semantic Web Platform · iPlant Semantic Web Program Foundational...

W O R K S H O P O N S E M A N T I C S I N G E O S PAT I A L A R C H I T E C T U R E S : A P P L I C AT I O N S A N D I M P L E M E N TAT I O N

O c t o b e r 2 0 1 3

D A M I A N G E S S L E R , P h . D .

S E M A N T I C W E B A R C H I T E C T

U N I V E R S I T Y O F A R I Z O N A

d g e s s l e r ( a t ) i p l a n t c o l l a b o r a t i v e ( d o t ) o r g

The iPlant Collaborative Semantic Web Platform

Classicism and Scholasticism

w w w . i p l a n t c o l l a b o r a t i v e . o r g 2

St. Thomas Aquinas Fra Bartolommeo (1472–1517)

Source: http://en.wikipedia.org/wiki/File:Thomas_Aquinas_by_Fra_Bartolommeo.jpg

Think, then look Look, then think

Empiricism


Source: http://en.wikipedia.org/wiki/File:Pourbus_Francis_Bacon.jpg

Look, then think Think, then look

Sir Francis Bacon Frans Pourbus the younger (1569–1622)

Scientific Method



Hypothesis

Prediction

Experiment Analysis

Conclusion

[ support or refutation ]

Sir Francis Bacon Frans Pourbus the younger (1569–1622)

The Experiment as the Rate Limiting Step


Charles Darwin Leonard Darwin, 1874

Source: http://en.wikipedia.org/wiki/File:1878_Darwin_photo_by_Leonard_from_Woodall_1884_-_cropped_grayed_partially_cleaned.jpg

Hypothesis

Prediction

Experiment Analysis

Conclusion


Bio(Geo)logy


Not too long ago: “If you love science and hate math, do Bio(Geo)logy”

All-in-one Analysis + Manuscript Prep + Data Management Plan


Not too long ago: “What? Me worry? My backups are safe”

Source: http://en.wikipedia.org/wiki/File:IBM_PC_5150.jpg

Quantity has a Quality all of its own


Hypothesis driven

Data Driven

Evidence-based Decision-making


http://blog.lib.umn.edu/ellis271/arch1701/bigstockphoto_Global_Warming_217540%203.jpg http://www.smartpower.org/blog/wp-content/photos/field_turbines.jpg

Decisions have downstream and unintended consequences; analyses and decisions about our Natural world that utilize a scientific approach bias our odds towards viable solutions.

The Analysis as the Rate Limiting Step


Source: http://en.wikipedia.org/wiki/File:Mapping_Reads.png

Hypothesis

Prediction

Experiment Analysis

Conclusion


high throughput sequencing

from climate change to

Re-revolutionizing the Revolution


13th century 20th century

Think, then look Look, then think

Hypothesis driven Data Driven

Re-revolutionizing the Revolution


16th century 21st century

Look, then think Think, then look

Data driven Hypothesis Driven

Hypoth

Predict

Exprmnt

Analysis

Conclsn



Hypoth

Predict

Exprmnt

Analysis

Conclsn

Hypoth

Predict

Exprmnt

Analysis

Conclsn

Hypoth

Predict

Exprmnt

Analysis

Conclsn

Analysis

Scientific Method, on hyper-cycles


Bridging HPC, Enterprise, and Web assets

High Performance Computing 500K core, 10 PetaFLOPS*

Petabyte scale storage

PetaFLOP: 1015 (million billion) floating point operations per second

Foundational Infrastructure iPlant Data Store, HPC, etc.

The iPlant Collaborative

Mission: To build a cyberinfrastructure for the nation’s plant scientists.

iPlant is a Service/Infrastructure project




Enterprise Class Discovery Environment, Atmosphere

Enterprise Class Virtual Work desk,

Cloud, Virtual machines

Discovery Environment: world class bioinformatics’ work station at your browser

Atmosphere: “instant” dedicated work station on the cloud: load, use, discard, repeat

breadth

dep

th





The Greatest Informatic Asset of all Time

Web

Just a portal and browser ... ... or an infrastructural asset?




Web

What is the infrastructural role for MODS, CODS, and trillions $$$ in web assets? How does iPlant engage, leverage, and enhance the Gramene’s, the TAIR’s, Soybase’s, SGN’s, MazieGDB’s, PlexDB’s, ..., of the world? How does iPlant engage anything that is not a downloadable, installable Linux/MS program?

MODS: Model Organism Databases CODS: Clade-Oriented Databases



iPlant Semantic Web Program



Web

SSWAP Semantic Integration

Semantic Pipelining

Enterprise Class Virtual Work desk

Cloud, Virtual machines

Distributed Semantic Web Services Logic-driven semantics





iPlant Semantic Web Program

Web

SSWAP Semantic Integration

Semantic Pipelining

Distributed Semantic Web Services Logic-driven semantics

It is the Semantic aspect of the Semantic Web that allows us to leverage the Web from being an external resource into an integrated infrastructural asset.

The Actors

You

Community MODS (Model Organism Databases)

and CODS (Clade Oriented Databases)

iPlant Computational Resources (e.g., TACC)

The World Your lab

The World

Semantic Mediation Layer


Semantic Mediation

Sidney Harris © 2006


The Antibody Analogy as the Mediation Layer


Simple Semantic Web Architecture and Protocol W3C OWL RDF/XML

• Establish the framework for Web resources to describe themselves and their offerings

• Establish the framework for ontological integration

• Engage first-order, description logic reasoning

• Provide a semantically enabled Discovery Server for service and pipeline coordination

http://sswap.info/protocol


http://sswap.info/protocol

SSWAP Enables Reasoner-assisted Workflows

• A protocol allows a reasoner to connect chains of resources (services) based on logical (not just lexical) matching of what various resources consume and produce.

• Web Discovery: from (any) Web site -> semantic pipeline

• Semantic Pipeline: single chain workflows of distributed semantic Web services hosted anywhere on the Web

• Workflows constructed via reasoner-assisted, first-order subsumption matching of the output of one service into the input of another


rdfs:subClassOf

Resource1 Subject Object



rdfs:subClassOf

iPlant Semantic Architecture

Data and Service Providers

Web Resource

Data

RDG RIG RRG

Algorithm

Web

Semantic Broker / Discovery Server

INTERPRETER

IND

IREC

TIO

N L

AYE

R

KB

BROKER

RDG RQG RRG EXPLICIT INTERFACE

IND

IREC

TIO

N L

AYE

R

Clients

RQG

Client

RIG RRG

Ontologies

Ontology Servers

Protocol Ontology

OWL RDF/XML

OWL RDF/XML

REPOSITORY

RESTful API

1

2

4 3

EXPLICIT INTERFACE

Semantic documents described in this talk: PDG: Provider Description Graph

RDG: Resource Description Graph RIG: Resource Invocation Graph RRG: Resource Response Graph

RQG: Resource Query Graph

PDG



sswap.info/example


From TreeGenes to High Performance Computing


Semantic Integration from Third-party Web sites DiversiTree

Javascript snippet to launch

data for Web Discovery

with the press of a button

HTTP API: sswap.info/api


JSON -> OWL RDF/XML

transformation

transparent to the user

(via the SSWAP HTTP API)


Web Discovery into Semantic Pipelines Reasoner-assisted Web workflows

Reasoner uses first-

order, description logic

to present services

and pipelines that can

operate on the data at

any given step

Just-In-Time Ontology Hosting sswap.info/jit



Web Discovery into Semantic Pipelines Reasoner-assisted Web workflows

Reasoner uses first-

order, description logic

to present services

and pipelines that can

operate on the data at

any given step


RESTful Pipeline Execution


Data Tree View “Ontologized” data and metadata

•on-the-fly Data Tree views

•pre-defined renderers


Semantic Integration into Third-party Web sites TreeGenes’ CartograTree

Third-party web sites can

engage as renderers on

result sets


TreeGenes’ Sequenced, Genotyped, and Phenotyped

Geographical browsing into

the TreeGenes database

• 1 265 tree species

• 901 113 sequences

• 24 142 786 genotypes

• 19 441 phenotypes

http://dendrome.ucdavis.edu/treegenes


AmeriFlux Sites CO2, Water, Energy

http://public.ornl.gov/ameriflux


WorldClim 1 Km2 climate grids

http://www.worldclim.org


ArcGIS Layers

Soil layers

http://maps2.arcgislonline.com


TRY-DB Phenotypes

http://www.try-db.org


CartograTree Custom user interface data selection and analysis

• Select

• Analyze

• Web Discovery


Semantic Integration from Third-party Web sites Data slicing and contextual augmentation


Direct and Indirect Data Referencing URI dereferencing of arbitrarily large data sets

Serialize the data itself,

or a URI to where the

data is located


High Performance Computing Services engage like any other Web services

Custom Service Parameterization



Publishing Pipelines Private data, shared service parameterization

Manage and Publish your Pipelines


Phylogenetics Pipeline runs are persisted in OWL and can start new pipelines


TreeViz Multi-location, multi-institution, Web/HPC run

Quick Vitals

• 185,000 lines of code

• 100+ libraries

• Open source

• Free SDK (Software Development Kit) for semantics and reasoning on your servers; run pipeline manager on our servers

• More info: sswap.info/wiki


http://sswap.info/wiki

Acknowledgements

Special thanks to:

• iPlant Collaborative

• UC Davis Dendrome / TreeGenes

• Semantic Web engineering by Clark and Parsia, LLC

• NSF grants #0943879 and #EF-0735191


The iPlant Collaborative Semantic Web Platform · iPlant Semantic Web Program Foundational...

Documents

Transcript of The iPlant Collaborative Semantic Web Platform · iPlant Semantic Web Program Foundational...