Semantic Web for Health Care and Life Sciences

Health Care and Life Sciences at W3C

Presented 19 December to CDISC to:

Pitch utility of the Semantic Web.
Motivate collaboration.

Eric Prud'hommeaux, Sanitation Engineer.
Last modified: $Date: 2009/07/13 23:34:23 $

This work is licensed under a Creative Commons Attribution 3.0 License, with attribution to W3C.

W3C Health Care and Life Sciences

A Story of Identifiers

What is the Semantic Web?

Who is HCLS?

Who is W3C?

What or Who is C27954

Ladies 9 carat filigree heart pendant on curb chain (C27954)

I intended the SDTM term for "ADOLESCENT".

Google suggests

MAFF_Rice cDNA Clone C27954
ACDE document C27954 (rural telecommunications)
A pretty necklace
...

But in a Controlled Vocabulary?

Less painful to resolve (go to authority).
But painful to extend (go to authority).

Classic Vocabulary Problem

You either

control term generation (at the authority).
control term interpretation (limit document use).

URLs are an Engineering Solution

Distributable term generation.
"Standardization" by popularity.
"Standardization" by authority.
Discovery by web (e.g. google).
Used in the Semantic Web.

Controled Vocabularies on the Web

Annotate them.
Dispute them.
Apply existing mapping languages,
e.g. SNOMED-CT to UMLS.

Many Mappings

Different use cases require different mappings.
Liberal mappings yield false-positives.
Conservative mappings yield false-negatives.
Need many well-described, individually-addressed mappings.

What is the Semantic Web?

Simple data model.
Suitable for arbitrary data expression.
Global identifiers and relations.
Built-in linking (e.g. HTML's href).
Ontology languages (RDFS, OWL, SKOS).

Blood Pressure Example

@prefix edns: <http://www.loa-cnr.it/ontologies/ExtendedDnS.owl#> .
@prefix foaf: <http://xmlns.com/foaf/0.1/> .
@prefix galen: <http://www.co-ode.org/ontologies/galen#> .
@prefix snomed: <http://termhost.example/SNOMED/> .

_:p   r:type galen:Patient ;
      foaf:family_name "Levin" ;
      foaf:firstName "Henry" .
_:c   edns:patient _:p ;
      edns:screeningBP _:scr .
_:scr dc:date "2006-03-18T18:23"^^xsd:dateTime ;
      edns:systolic "132"^^edns:mmHg ;
      edns:diastolic "86"^^edns:mmHg ;
      edns:posture snomed:_163035008 . # SNOMED:sitting

Groups/Sets/Classes

[Named] collections of things with similar attributes.

Groups/Sets/Classes

[Named] collections of things with similar attributes.

Groups/Sets/Classes

[Named] collections of things with similar attributes.

OWL Expressivity

Allows you to define classes.

(a,b,c) set enumeration

union

disjunction

algebraics

intersection

complement

restriction

cardinality

equivalence

Semantic Web Scalability

Decentralized contributions to unambiguous (read "global") vocabularies.
Provides tools to stitch together with rest of SemWeb for best data re-use.
Makes simple stuff (assert, query, merge data) easy.
Makes hard stuff (complex rules/OWL) possible.

Who is HCLS?

W3C Semantic Web Health Care and Life Sciences Interest Group.

90+ members from industry and academia.

1 KB (350m+ triples)

2 papers

Ontology mapping code.

Current Work

BioRDF – integrated neuroscience knowledge base
Task lead - Kei Cheung
Linking Open Drug Data – aggregation of Web-based drug data
Task lead - Chris Bizer
Clinical Observations Interoperability – patient recruitment
Task lead - Vipul Kashyap
Scientific Discourse – building communities through networking
Task leads - Tim Clark, John Breslin
Terminology – SemWeb representation of existing resources
Task lead - John Madden

Other Projects: Clinical Decision Support, URI Workshop...

HCLS KB

A Knowledgebase for Neuroscience

Built by Neurocommons and the HCLS BioRDF task force.

Precise answers to complex questions:

What proteins are associated with pyramidal neurons
AND are involved in signal transduction.

Integrate Databases

Mesh
Pubmed
Entrez Gene
Gene Ontology
...

Real Query

Linking Open Drug Data

Clinical Observations Interoperability

Bridge relational databases to HL7/RIM.
Bridge HL7/RIM to SDTM.
Encode some clinical protocol (inclusion/exclusion criteria) in SDTM.
Use drug ontology to map drug classes to brands.
Protocol selection in SDTM translated to RIM, relational query.
Compare this SemWeb-y approach to other approaches.

Scientific Discourse

Terminology

Importing terms into Semantic Web.

Start with pharmacological classifications in SNOMED-CT and UMLS
Create URL identifiers for terms.
Express term relationships (e.g. pre-coordinated terms) in OWL.
Document process and and expressivity of schema languages.

W3C from 10,000 feet

International industry consortium.

Effective standards.

Technology development.

Web architecture oversight.

Effective Standards

Working/Interest Groups.

Clear IP: (RF or RaND).

Public review.

Documented process.

Architectural input from many experienced members.

Interoperability on a web scale.

Next Steps

Help CDISC consider SemWeb approach.

Give early access to use cases and best practices.
Listen to needs.
Gain more domain expertise.
Share work-load.

Done

Technical slides follow for those who are interested in

OWL
RDF
COI details
RDF compared to XML

Explain RDF

BP slide.
nodes and arcs picture
RDF/XML
sparql
RDF stores (triple and other)
mappings (à la COI)

Simple statement

_:p   r:type galen:Patient .

Add a property ...

_:p   r:type galen:Patient ;
      foaf:familyName "Levin" .

... and another.

_:p   r:type galen:Patient ;
      foaf:familyName "Levin" ;
      foaf:givenName "Henry" .

Statements about ...

_:p   r:type galen:Patient ;
      foaf:familyName "Levin" ;
      foaf:givenName "Henry" .
_:scr dc:date "2006-03-18T18:23"^^xsd:dateTime .

... other objects ...

_:p   r:type galen:Patient ;
      foaf:familyName "Levin" ;
      foaf:givenName "Henry" .
_:scr dc:date "2006-03-18T18:23"^^xsd:dateTime ;
      edns:systolic "132"^^edns:mmHg .

... connect in the fabric.

_:p   r:type galen:Patient ;
      foaf:familyName "Levin" ;
      foaf:givenName "Henry" .
_:c   edns:patient _:p ;
      edns:screeningBP _:scr .
_:scr dc:date "2006-03-18T18:23"^^xsd:dateTime ;
      edns:systolic "132"^^edns:mmHg ;
      edns:diastolic "86"^^edns:mmHg ;
      edns:posture snomed:_163035008 .

Algebraic Properties

transitiveProperty:
- :Bob :worksFor :Sue .
- :Sue :worksFor :Tom .
- :worksFor a owl:TransitiveProperty .
- ⇒ :Bob :worksFor :Tom .
inverseProperty:
- :hasChild owl:inverseOf :hasParent .
reflexiveProperty:
- :friendOf a owl:SymmetricProperty.

Uses

Type inference.
- :Bob a :MGHcharlesStOncologist .
Identity inference.
- merge the :Bob and :BobSmith nodes.
Consitency rules.
- An :MGHcharlstStPatient
  - is an :MGHpatient.
  - has a :physician of type (:MGHcharlesStOncologist or :MGHcharlesStOptician or ...).
  - has an MGH Charles Street consent form.

OWL type inference

onProperty
- :fluffy :hasChild :fluffyPrime .
- :fluffyPrime a :kitten.
- ⇒ :fluffy a :cat .

OWL identity inference

functionalPropery
- :Bob :hasChild :Sue .
- :BobSmith :hasChild :sue .
- :hasChild a owl:FunctionalProperty .
- ⇒ :Bob is the same as :BobSmith
InverseFunctionalPropery
- :Bob foaf:mbox <mailto:bobS@foo.example> .
- :BobSmith foaf:mbox <mailto:bobS@foo.example> .
- foaf:mbox a owl:InverseFunctionalProperty .
- ⇒ :Bob is the same as :BobSmith

OWL restriction expressivity

PARTIAL — This restriction is a necessary condition for this class.
COMPLETE — This restriction completely defines the extent of this class.
AND — This restriction requires both things to be true.
OR — This restriction requires either thing to be true.
ALL — The values of some property must all be in some set.
SOME — Some of the values of a property must be in some set.