OntologicalModelandApproaches intheIntegratedBiomedical - - PowerPoint PPT Presentation

▶
ontological model and approaches in the integrated
SMART_READER_LITE
LIVE PREVIEW

OntologicalModelandApproaches intheIntegratedBiomedical - - PowerPoint PPT Presentation

Feb 02, 2024 376 likes •1k views

OntologicalModelandApproaches intheIntegratedBiomedical DatabaseProject (IBMD) HiroshiTanaka TokyoMedicalandDentalUniversity Whatis theIntegratedBiomedicalDB project? project?

slide-1
SLIDE 1

OntologicalModelandApproaches intheIntegratedBiomedical DatabaseProject (IBMD)

HiroshiTanaka TokyoMedicalandDentalUniversity

slide-2
SLIDE 2

Whatis theIntegratedBiomedicalDB project? project?

slide-3
SLIDE 3

IntegratedBiomedicalDBProjects

  • Governmentcommissionedproject

– Startedat2007inTokyoMedical&DentalUniv.

  • byMEXT(MinistryofEducation,Culture,Sports,

ScienceandTechnology) ScienceandTechnology)

  • AsubprojectoftheNationalProjectsforDBs

integrationinLifescience fields

  • Maingoalofthisprojectis

integrationofdiseasedatabasesinJapan

slide-4
SLIDE 4

BiomedicalOntologyProjects inJapan

  • Two governmentcommissionedmedical
  • ntologyprojectsinJapan
  • Ours:ScientificDBintegrationbyMEXT
  • ClinicalInformationSystembyMHLW

(MinistryofHealth,LabourandWelfare) “JapanMedicalOntologyDevelopmentProject forAdvancedClinicalInformationSystem”

– Dr.Imai’stalk

  • Goodcollaborationbetweenbothprojects
slide-5
SLIDE 5

NationalProjectforDBintegrationin LifeScienceField(MEXT)

  • IntegratedDBProjectinLifeScience

– Startedfrom2006

Background

  • DBsinLSinJapanwerescatteredinvariousinstitutes

– DNAsequence(DDBJ)inNIGinMishima – DNAsequence(DDBJ)inNIGinMishima – ProteinDB(PDBJ)inOsakaUniversity – PathwayDB(KEGG)inKyotoUniversity – UnlikeNCBIinUSandEBIinEurope

Goal

  • EstablishNationalCenterforDBintegrationinLS

(DBCLS)

– developintegratedDBservice – commonportalsforDBsinLS

slide-6
SLIDE 6
slide-7
SLIDE 7
slide-8
SLIDE 8
slide-9
SLIDE 9

AllocationOrganizationfor IntegratedDBProjects

  • TokyoMedicalandDentalUniversity(TMDU),

InformationCenterforMedicalScience

– commissionedbyMEXT – tointegratescattereddatabasesonvariousdiseaseinJapan – tobeaNationalCenterforDiseaseDatabaseIntegration – tobeaNationalCenterforDiseaseDatabaseIntegration – from2007,IntegratedBiomedical(clinical)DB

  • KyotoUniversityBioinformaticsCenter

– IntegrationofDrugInformation

  • TokyoUniversityDept.HumanGenetics

– IntegrationofGeneticPolymorphism

slide-10
SLIDE 10

Challengesfor IntegratedBiomedicalDBprojects

  • Developmentof OntologybasedIntegrative

InterfaceforDiseaseDBsearch

  • IntegrationoftheCoreClinicalDBsinJapan

– ParkinsonDiseaseDBinOsakaUniv. – ParkinsonDiseaseDBinOsakaUniv. – GEMDBJinNationalCancerCenter – Soon

  • EstablishEthicalCodeforPublicizingClinical

CaseDB

slide-11
SLIDE 11

Then,howcanwedescribe Disease? Isitphysicalentity,or Isitphysicalentity,or justconceptual ?

slide-12
SLIDE 12

DiseaseModelinginDatabase

  • Whatistheontologicalcharacteristics
  • fdisease?
  • Complexityof“Disease”
  • Diseaseisamultifaceted,

DiagnosticTherapeutic Prognosis Diseaseprogression epidemiological populational beha social

Dise

multilayeredentity

– moleculargenetic – tissueorganic – individual – diagnostictherapeuticprognostic – behavioral(medicalpractice) – populationalepidemiological

moleculargenetic Organsystems Individualobservationa DiagnosticTherapeutic biological behavioral

Disease

tissueorganic

slide-13
SLIDE 13

DiseaseModelinginDatabase

  • Complexityofthe“Disease”

– Multifaceted,multilayeredentity – Incompletenessofdiseaseknowledge – Multiplicityofrelatedsciences – Multiplicityofrelatedsciences

  • biological,psycological,behavioral,conceptual,

socialscience

  • Mainoppositestandpoints

– PhysicalvsConceptual – CausativevsObservational

slide-14
SLIDE 14

Howdowedescribedisease

  • FormalDescriptionofDisease
  • DiseaseViewarenowchangingsince

RevolutionofMolecularmedicine RevolutionofMolecularmedicine

  • ConventionalView

– MultilayeredPhenotypicalDescription – EssentiallyObservational – (Place,Organ)X(Pathomorphology)

  • myocardial(place)infarction(pathology)
slide-15
SLIDE 15

RecentChangesofDiseaseView

  • AdvancesinMolecularMedicine

– diseasegenetics

  • diseasecausative(related)gene
  • geneticpolymorphism(SNP,ms)

SNPs

  • geneticpolymorphism(SNP,ms)

– diseaseomics

  • geneticexpressionprofile
  • proteome,metabolome

– diseasemolecularpathway

  • distortedsignalpathwayor

regulatorynetwork

DNAmicroarray SNPs

slide-16
SLIDE 16

IndividualLevel IndividualLevel

Top&dow Bottom

  • Diseaseishierarchicallyorganized

“distortedmolecularnetwork”

Tissue Tissue& &OrganLevel OrganLevel CellularnetworkLevel CellularnetworkLevel

downCausality

  • ttom&upCausality

!

  • !"
  • !"
slide-17
SLIDE 17

Ontology

  • Formalrepresentationofasetofconcepts

withinadomainandtherelationshipsbetween thoseconcepts

– usedtoreasonaboutthepropertiesofthatdomain, andtodefinethedomain andtodefinethedomain – “Formal,explicitspecificationofashared conceptualisation“(Gruber,klsstanford)

  • Controlled(Formal)Vocabulary

– usedtomodeladomainforknowledgesharingand reuse – thetypeofobjects/conceptsthatexist,andtheir propertiesandrelations

slide-18
SLIDE 18

ChallengesinBiomedicalOntology

Ontologymismatchbetween clinicalthinkingandOmicsmedicine

  • Eachinformationsystem

hasitsontologyasa basis

Information system

Information system

basis

  • Mismatchbetween

Clinicalontologyand Bio&ontology

terminology

  • ntology

terminology

  • ntology

Clinical Ontology Bio&ontology

slide-19
SLIDE 19

MismatchofThinking

  • Clinicalthinking

– Organsanddiseasesareunitsofconcepts – Clinicalphenotypical;diseaseisdefinedonpathological, morphological(changes)base – Essentiallygoal&oriented:diseasecaredirected – Essentiallygoal&oriented:diseasecaredirected – Topdown

  • Molecular(Omics)medicinethinking

– Molecularfunctionandtheirfunctionalrelationtoother molecules – Productsofgeneexpressionareunitsofconcepts – Bottomup

slide-20
SLIDE 20

! Clinicalthinkingand“Omicspace”

symptomatology symptomatology

Clinicalinformation Clinicalinformation

Pathway Pathway signalome signalome networkome networkome

Omicspace Omicspace

#

  • Pathophsiology

Pathophsiology Etiology Etiology Genome Genome SNP,Haplotype SNP,Haplotype Transcriptome Transcriptome Proteome Proteome metabolome metabolome

slide-21
SLIDE 21

ExistingClinicalandBio&Ontology

  • Clinicalontology

– Semanticnetwork,UMLS,Galen(SNOMED)

  • GeneOntology

– Molecularfunction,process,cellularlocationof geneproducts – Nowonly,Eukaryotes

slide-22
SLIDE 22

SeveralprojectsforIntegrationof Bio/ClinicalOntology

  • OpenBiomedicalOntologies(OBO)

– NationalCenterforBiomedicalOntology(Mussen) – OBOFoundry(Smith)

  • OtherProjects
  • OtherProjects

– OntologyforBiomedicalInvestigations – ULMSplanstoinvolve GeneOntology – DiseaseOntology – Soon

slide-23
SLIDE 23

IntegrativeClinicoOmicOntology (possibletransitform)

  • Globalstructurefollows

theframeworkof clinicalontology

  • Withintheframework

$ %!"

  • Withintheframework

bio&ontologyis employedtoprovide bottom&uprelationof themeaningof phenotypicalentities

molecular manifes& tations (makers) therapy (drugs) molecular

slide-24
SLIDE 24

PracticalSolution

  • ClinicalNosologicalOntology

– CoventionalTextbookknowledgeOntology withpatientmedicalinformationontology – usedforOntologyforDiseaseDBs – usedforOntologyforDiseaseDBs integration

  • MultilayeredClinicalOmicsOntology

– Stillunderdevelopment – Butwithlinkedmultilayereddataschema – integratedClinicalOmicsDatabase(iCOD)

slide-25
SLIDE 25

SeveralResultsofIBMDBproject

slide-26
SLIDE 26

ResultsforOngiongIBMDBproject

  • DevelopedthefirstversionofNosological

IBMBDOntology

  • MutlilayeredIntegrativeClinicalOmicsDB
  • MutlilayeredIntegrativeClinicalOmicsDB
  • DevelopedOntologybasedInterDB

SearchSystem

  • TrialsystemforintegrativeDBsearch

betweenParkinsonDiseaseDBandiCOD

slide-27
SLIDE 27

IBMDBOntology

slide-28
SLIDE 28

Disease Clincial

Classification

Systems Omics Pathology

Ontol

Concep Concep IntegrationLevel IntegrationLevel

EstablishedKnowledge (textbook) EHR CENen13606

Archetype

2linesofintegrationapproach

Nosology3based integration Systems3pathobiology3 basedintegration

MetaClassification

Nosologicalclassification SystemsPathoBiological classification

Integrationbasedon establishedconcept classification

&'() ３

３ ３ ３ Levelmodelfordatabaseintegration

Disease Clinical

Cancer Data DataStructure Terminology NeurologicalDisorder

Systems Omics Pathology

Ontology Info.Model

  • nceptLevel
  • nceptLevel

Info.ModelLevel Info.ModelLevel Concept Library Reference Model

interlock

HL7

SNOMEDCT

2Level BIMS CommonTemplate Data DataStructure Data DataStructure

slide-29
SLIDE 29

DiseaseTerminolologyandClassification

CondensedCross ClinicalOntology DiseaseTerminology DiseaseClassification

Nakaya,J.,Sasaki,K.,andTanaka,H.(2006)CondensedCrossClinicalKnowledge,ComputerScience,IJCSNS.6(7A).6&11.

Thisisthediseaseterminologyandclassification. Diseasesareclassifiedwithcombinatedanatomicalhierarchyandetiologicalhierarchy. AconceptualunitofaDiseasseisdescribedwiththe3rdnormalizedskeletontemplatewhichcanbe calledasacontentmodel.

#* ＋ +*

, ,)

  • ./"

, #

slide-30
SLIDE 30
  • /

# 0 ,

  • $

+ +

  • /

, " * ',* %* +0%

  • 1
  • 2#

, , $

  • /
  • ,
  • /

, /( , $ ,

3 1( '!(

2# 1/ 3 $ /% % %

  • 45

,, , )

,*

slide-31
SLIDE 31

NosologicalOntology

（anexampleofCompositeindexofanatomyandetiology)

slide-32
SLIDE 32

AnexampleofLiverdiseaseindex

slide-33
SLIDE 33

ApartofParkinsonDisease

slide-34
SLIDE 34

integratedClinialOmicsDB (iCOD) (iCOD)

slide-35
SLIDE 35
  • Integratethemolecularomicsinformationand

clinicalandpathodological,lifestyleinformation

  • Comprehensivedatabasebasedontheconceptof

“omics&basedsystemsmedicine”

$ $

Purpose

$ $

  • $

$

  • Data

Data Mining Mining $ $ ', ',

slide-36
SLIDE 36

Cases

cases

  • mics information

transcriptome CNV specimen normal specimen

  • rmal

Hepatocellular carcinome １９３ １９３ １９３ １９３ １５２ １５２ １５２ １５２ ９６ ９６ ９６ ９６ １０２ １０２ １０２ １０２ ３５ ３５ ３５ ３５ stored ４１ ２９ ０ ３４ ０ fresh １３４ １０５ ８１ ６６ ３５ metastasis １８ １８ １５ ０ ０ metastasis １８ １８ １５ ０ ０ colon cancer １８４ １８４ １８４ １８４ １３１ １３１ １３１ １３１ ２８ ２８ ２８ ２８ ３９ ３９ ３９ ３９ ４０ ４０ ４０ ４０ colon １２８ １０２ ２８ ３６ ３６ rect ３７ ２９ ０ ３ ４

  • ral tumor

１４８ １４８ １４８ １４８ ２０ ２０ ２０ ２０ ０ ０ ０ ０ ６４ ６４ ６４ ６４ ２ ２ ２ ２ stored ６４ ０ ０ ６４ ２ fresh ８４ ２０ ０ ０ ０ total ５２５ ５２５ ５２５ ５２５ ３０３ ３０３ ３０３ ３０３ １２４ １２４ １２４ １２４ ２０５ ２０５ ２０５ ２０５ ７７ ７７ ７７ ７７

slide-37
SLIDE 37

iCOD:integratedClinicalOmicsDatabase

Toppage CaseArchvie CaseList

//

CaseDetails

(!

slide-38
SLIDE 38

Toppage

slide-39
SLIDE 39

CaseList

//

slide-40
SLIDE 40

CaseDetails(1)

4$5 6!

slide-41
SLIDE 41

CaseDetails(2)

slide-42
SLIDE 42

ClinicalOmicsAnalysis

slide-43
SLIDE 43

Clinical3&layeredOmicsMap

slide-44
SLIDE 44

Omics dataanalysis

'

Pathwaymap OmicDownload PosMed Legenda

  • TranscriptomemappedonKEGG
slide-45
SLIDE 45

SemanticNavigationSystem

Weshowthesemanticnavigationsystem whichwecansayasacontentinterfacebasedonthecontentmodel.

slide-46
SLIDE 46

IntegratedDBGuidingSystem

  • forCoreDB （2DBs＋

＋ ＋ ＋α（ （ （ （10DBs） ） ） ） ： Nowininvestigation) WearedevelopingSearchingSystembasedontheontology

  • UserI/FManagerconvertuserinputtostandardizedwordwithterminology

andthesaurus.

  • Navigatingenginenavigatesuserstotargeteddatabasessemantically.
  • DataformatmapperabsorbsformatdifferencesofDBs.

DBsearch

DataFormat1 DataFormatn

Thecontentmodelisthebasic templateofbothontologyanddata formatinthesystem ユーザー User I/F Manager

DB Manager Navigating Engine

Nosological Ontology Terminology Thesaurus DataFormat Mapper

DataFormat2 DataFormat3

Standardized DataFormat

Data Formatn Data Formatn Data Formatn Data Formatn DataFormatn

ContentModel

slide-47
SLIDE 47

PrototypeDemo

47

PPTversion

slide-48
SLIDE 48

IBMDTopPage

http://ibmd.tmd.ac.jp

48

slide-49
SLIDE 49

InputBoxforSearch

DBMapofIBMD

  • nDiseaseClassification

SearchWindow

49

ResultsofSearching

slide-50
SLIDE 50

Example

– Inquiry“Symptom Symptom： ： ： ： ： ： ： ：Depressivestate” Depressivestate”

  • CancerＤＢ inTMD

– ＜Ｄ＞ ０ cases – ＜Ｓ＞ byThesaurus,semantictransform “Depressivestate→Anorexia” Depressivestate→Anorexia” ⇒ ２ ２ ２ ２ ２ ２ ２ ２ cases cases – ＜Ｓ＞ “Depressivestate→Lassitude” Depressivestate→Lassitude” ⇒ ４ ４ ４ ４ ４ ４ ４ ４ Cases Cases » onecaseisdoubled5cases

  • ParkinsonDiseaseＤＢ inOsakaUniv.

– ＜Ｄ＞０ cases – ＜Ｓ＞Ontology、” ” ” ” ” ” ” ”Depressivestate→slowmovement” Depressivestate→slowmovement” – – ⇒ ６ ６ ６ ６ ６ ６ ６ ６ cases cases

slide-51
SLIDE 51

Input

１．Category:”Symtoms Symtoms”、SearchInputWord:”Depressivestate Depressivestate”

51

２．Push SearchButton

slide-52
SLIDE 52

Results

52

１．FromTMD TMDＤＢ ＤＢ ＤＢ ＤＢ ＤＢ ＤＢ ＤＢ ＤＢ <Direct>Ocases

<Semantic>「Symptom:Depressivestate→Anorexia Anorexia」 2 2cases <Semantic>「Symptom:Depressivestate→Lassitude Lassitude」 4 4cases Duplicatecase1case→ （Total5cases Total5cases）

２．FromOsakaUniv. OsakaUniv.ＤＢ ＤＢ ＤＢ ＤＢ ＤＢ ＤＢ ＤＢ ＤＢ <Direct>0cases

<Semantic>Symptom:Depressive→SlowMovement Symptom:Depressive→SlowMovement 6cases

slide-53
SLIDE 53

DetailedData

Clickherefordetaileddata

53

slide-54
SLIDE 54

DetailedDatasubWindow

HittedPatientInformation isdisplayed accordingpatientcontent Thispanelisbasedon thearchetypetemplate(thecontentmodel)ofdisease

54

accordingpatientcontent model(template)

slide-55
SLIDE 55

PatientContent

Ex：PatientHistory subwindow

55

slide-56
SLIDE 56

ICD11andHIM3TAG

slide-57
SLIDE 57

TAG3HIM presentstatus

  • 1.

ContentModelGroup – TreattheoriginalusecaseofICD11(Stefany)

  • 2.Informationgroup

– AlanRector(Univ.ofManchester,UK),John,Chris,SCT

  • 3.

ContentModelfrontendforeachTAG, – CategorialStructureforRaredisease

  • JeanMariewithRareDiseaseTAG

– IBMDBmodelismodifiedforinternalmedicine

  • JunNakayaandHiroshiTanakawithInternalMedicineTAG

– Robertwilltakecareofthesethings

  • 4.

SCT(SNOMED)Coordination – IHTSDOharmonizationpanelwilltakethisissue.

  • (Kent,AlanRector(Univ.ofManchester,UK),Chris,Olivier)

– GOandotherontologieswillcovertheremainedarea.

slide-58
SLIDE 58

DataExchangeFormat （asanInfo.Model）

slide-59
SLIDE 59

/76"' /76"' 4/6'5 4/6'5 /$89:;<:=Passed/

US,UK,Canada,Korea,Italy,Israel,Australia,Japan

LedbyJunNakaya

Nakaya,J.,Hiroi,K.,Yang,W.,Ido,K.,Kimura,M.(2006)"GenomicSequenceVariationMarkupLanguage (GSVML)forGlobalInteroperabilityofClinicalGenomicsData(#!)".AsiaPacificAssociationfor MedicalInformatics2006Proceedings.A01.1&8. ; JunNakaya,MichioKimura,HiroshiTanaka

slide-60
SLIDE 60

OutlinedStructureofGSVML

TheGSVMLhashierarchicalstructure. TheentrypointofGSVMLisgenomicsequencevariation.

Nakaya,J.,Hiroi,K.,Yang,W.,Ido,K.,Kimura,M. (2006)(BestPaperAward)APAMI2006A01:1&8.

TheGSVMLhas3Datacriteriaasvariationdata,directannotation,andindirectannotation. Thesecriteriahavetheinternalrelationsmainlybasedonthestatitics.

slide-61
SLIDE 61
  • HiroshiTanaka

(Director) TMDUBiomedicalOntologyGroup JunNakaya(Leader) JunNakaya(Leader) KeisukeIdo KaeiHiroi