CDISC on EVS Sept05

1

Overview for NCI Enterprise Vocabulary Services
(EVS) for the caBIG Integrative Cancer Research
Workspace 11/09/05
Sherri de Coronado, MS, MBA Frank Hartel,
PhD Margaret Haber, RN, OCN Enterprise
Vocabulary Services National Cancer Institute
2
Outline

• Terminology goals and semantic integration
• NCI Enterprise Vocabulary Services
• NCI Thesaurus (NCIt)
• NCI Metathesaurus
• API Access through caCore
• Thesaurus and Metathesaurus Production Process
  and Curation

3
Goal Controlled Terminology for Semantic
Interoperability

• Systems cannot exchange or use information if
  they use incompatible codes or tokens to signify
  meaning
• Terminology services provide token and codes
• Proper use of them assures consistent meaning
  across the enterprise

4
Supporting CORE Infrastructure

• Enterprise Vocabulary Services (EVS)
• Core Semantics for caCORE and many other
  applications
• Public access browsers
• APIs
• cancer Data Standards Repository (caDSR)
• ISO 11179 compliant metadata repository
• Common Data Elements (CDEs) for multiple
  templates, such as Case Report Forms, linked to
  EVS terminology
• cancer Bioinformatics Infrastructure Objects
  (caBIO)
• UML Models annotated with EVS concepts/terms,
  loadable into caDSR
• Public access APIs

5
Binding Data, Metadata to Terminology - caCORE SDK

• UML Modeling Tool (provided by user)
• Information model that will define data classes,
  attributes and relationships
• Semantic Connector
• Annotate UML model with ontology concepts
  bridges the world of databases to that of
  structured semantics.
• UML Loader
• Loads model into the caDSR metadata registry
• Model and associated semantics are available at
  runtime
• Code Generator
• Model and a code template are inputs into
  generator
• Creates the caCORE-like n-tier software system
  with Java and Web Services APIs

6
How EVS used with caDSR

From D.Warzel
7
caGrid Data Description Infrastructure Includes
EVS

• Objects defined in UML, converted into ISO/IEC
  11179 Administered Components, in turn registered
  in the caDSR
• Object definitions draw from vocabulary
  registered in the EVS, and their relationships
  are thus semantically described

8
Enterprise Vocabulary Services

• Services and resources that address NCI's needs
  for controlled vocabulary http//www.nci.nih.gov/
  EVS
• A collaboration between
• NCI Office of Communications
• Physician Data Query (PDQ), Cancer Information
  Service, the NCI web portal www.cancer.gov, and
  Drug dictionary
• NCI Center for Bioinformatics
• Bioinformatics Core Infrastructure (caCORE),
  including metadata repository (caDSR) and object
  models built using EVS terminology for core
  semantics

9
EVS Products

• Clinical, translational, and basic research
  terminology have overlapping but specialized
  needs, therefore EVS assists to
• Integrate different conceptual frameworks
• Create terminological and taxonomic conventions
  across systems
• Vocabulary Products
• NCI Thesaurus an ontology-like DL terminology
• NCI Metathesaurus based on UMLS, maps
  vocabularies
• External vocabularies maintained and served
  MedDRA, HL7, NDF-RT, LOINC, etc.

10
In caBIG, NCI EVS Goal

• Support vCDE efforts to deploy clinical,
  translational, and basic research terminology
• Assure that widely used terminological and
  taxonomic conventions are followed
• Support Workspace terminology needs
• Help find and use relevant existing terminology
• Incorporate needed terminology in EVS
• Provide liaison between NCI and caBIG Workspaces

11
NCI Thesaurus (NCIt)

• Reference Terminology for NCI, Partners
• A Federal Standard Terminology
• Broad coverage of the cancer research and
  clinical domain including prevention and
  treatment trials
• Neoplastic and other Diseases
• Findings and Abnormalities
• Anatomy, Tissues, Subcellular Structures
• Agents, Drugs, Chemicals, Combo Chemo
• Genes, Gene Products, Biological Processes
• Animal Models Mouse, other
• Research techniques and management, apparatus,
  clinical and lab, radiology, imaging

12
NCI Thesaurus (2)

• Published Monthly
• Public domain, open content license
• Available on-line and by download (OWL, Ontylog
  XML, flat files)
• 48,000 Concepts hierarchically organized into
  trees
• Description-logic based
• Roles establish machine readable semantic
  relationships between Concepts, ex.
• Carcinoma Disease_Associated_with_Disease
  Lytic Bone Lesions,
• TP53 Gene_associated_with_Disease Breast
  Carcinoma

13

• NCI Thesaurus is Deployed
• http//nciterms.nci.nih.gov
• http//www.nci.nih.gov/EVS
• (full documentation)
• API caCORE public access
• Fulfills NCI and collaborators needs for
  controlled vocabulary
• Public domain, open content license

14
(No Transcript)
15
Disease, Disorder or Finding (multiple
inheritance)
Disease, Disorder (7000 concepts)
Organized by Site or by Type
Finding (Cancer TNM Finding, Morphologic Finding,
Lesion, Clinical Course of Disease, etc.)
16
Concept History
Draw Graph
Roles
Termtypes and Source
17
Concept History Stage_IV_T-Cell_Non-Hodgkin_s_Lymp
homa (code C8668)
Merge or Split would result in reference code
18
Graph Selected Roles, distance n2
19
Links to other sources
20
(No Transcript)
21
Listserv
Download
22
NCI Metathesaurus

• Filtered UMLS Metathesaurus extended with
  additional required vocabularies
• 930,000 concepts, 2,200,000 terms and phrases
  with definitions
• Mappings among over 50 vocabularies
• Extensive synonymy Over 40,000 terms for
  neoplasms mapped to 7,000 concepts
• Used as online dictionary and thesaurus, for
  mapping and document indexing

23
NCI Metathesaurus (2)

• Minor releases monthly, Major releases twice a
  year. (Soon to be updated monthly with NCI
  Thesaurus)
• Provides a mapped overlap and partial
  inter-relation of current versions of NCI and
  partner required vocabularies, ex. The ICDs,
  MedDRA, SNOMED, MeSH, HCPCS (procedures), LOINC
  (lab values), drug terminologies (VA NDF-RT, AOD,
  RxNORM, Multum, NCI Thesaurus drugs, etc.)

24
(No Transcript)
25
http//ncimeta.nci.nih.gov
NCI added sources in red
26
Advanced Search
Search by CUI or Code
Search by Source
27
EVS Products Services Are Open

• NCI Thesaurus is Open Content ftp//ftp1.nci.nih.g
  ov/pub/cacore/EVS/ThesaurusTermsofUse.htm
• NCI Metathesaurus is Mostly Open Source
• See Each Sources License http//ncimeta.nci.nih.
  gov/MetaServlet/GenerateSourcesServlet
• NCI EVS Servers Are Freely Accessible
• On the Web
• Via API
• All Software Developed by NCI EVS is Public Open
  Source and Free for the Asking

http//nciterms.nci.nih.gov and
http//ncimeta.nci.nih.gov
http//ncicb.nci.nih.gov/core/caBIO
http//ncicb.nci.nih.gov/core
28
API Access to EVS
Info on how to get concept information from
either NCIt or NCI Meta avail in caCORE 2.0 User
Guide
29
Access to NCI Thesaurus via APIs
30
Notional Java Code Assuming a property called
FDA_Table and our wanting to find the table
contents for the Dosage Form table, DRG-00201

• Gilbertos example code would go here

31
EVS Curation

• NCI Thesaurus
• Domain editors
• Two internal baselines per month
• One baseline for publication monthly with history
  file
• Placed on test server
• Pre-Thesaurus as well
• Editing in progress to fix errors, meet end user
  needs, support caDSR/ UML model annotation

• NCI Metathesaurus
• Moving to monthly updates with NCI Thesaurus in
  MEME
• Insertion/ updates of additional sources as
  needed, including UMLS and special sources like
  MedDRA
• Matching and Merging/ QC processes/ production to
  Meta test server, promoted to production

32
TerminologyDevelopment Guidelines

• Develop a content model for a domain
• Leverage existing sources where appropriate
• (VA NDF-RT, RxNorm, LOINC, etc. )
• Develop unique content where needed
• (Cancer genes and diagnoses, drugs and therapies,
  molecular abnormalities, clinical trial standard
  terminology etc.)
• Link to other information sources and standards
  using URLs as possible
• (GO, Swissprot, drug formularies, trial
  protocols)
• Federate, merge or map with other standard
  terminology for semantic integration

33
EVS Collaborations

• Many Active Collaborations
• Federal FDA, VA, CDC, and Various NIH Institutes
  such as NHLBI, NIDCR
• Major Standards Organizations HL7, CDISC, W3C,
  FHA
• Cancer Centers and Cancer Cooperative Groups
  (caBIG, caGRID)
• Numerous Research collaborators such as the
  Microarray Gene Expression Data Society (MGED
  Ontology, FuGO)

34
ContactsSherri de Coronado, MS, MBANCI
Center for Bioinformaticsdecorons_at_mail.nih.govF
rank Hartel, PhDNCI Center for
Bioinformaticshartel_at_mail.nih.govMargaret
Haber, RN, OCNNCI Office of Communicationsmhaber
_at_mail.nih.gov