利用混合模型研究基因與疾病的關(guān)聯(lián).pdf

1、東北師范大學(xué)碩士學(xué)位論文利用混合模型研究基因與疾病的關(guān)聯(lián)姓名：車平申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：概率論與數(shù)理統(tǒng)計(jì)指導(dǎo)教師：史寧中20040501AbstractThecase—controlstudyisatooloftenusedinepidemiologyHowever，ifthereiSaconfouderwhichiSassociatedwithboththecageandtheriskfactorstudied，confoundin

2、gwillbebroughtintothestudyanditwillcometoanincorrectconclusionofassociationThemainapproachtosolvethisproblemistostratifythepopulationstudiedthroughtakingtheconfounderasthestratifyingvariableInmostcases，theCOIlfounderisun

3、observable，thuschoosingapropercovariableisveryimportantInthispaperwetakethemultilocusgenotypedataasthecovariabletotheconfounderandproposeanapproachusingmixturemodelstodetecttheassociationbetweenacandidategeneandadisease，

4、whilethereissubstructureinthepopulationinacase—controlstudyInourapproach，wemakefuUuseoftheknowninformationandstratifythepopulationusinggeneticmarkerswhileconsideringthatHardy—WeinbergequilibriumholdsineachsubpopulationEM

5、algorithmisusedinthecourseofestimatingMeanwhile，theassociationbetweenthecandidategeneandthediseaseisestimatedAdditionally，wehavetestedourapproachusingsimulateddatabasedonallelefrequenciesin12shorttandemrepeatfSTRllOCifro

6、mfoursubpopulations，andcomparedourresultswiththatofSatteneta1(2001)ItcanbeconcludedfromthesimulationthatourapproachiSmoreaccurateandmoreefficientBesides，weconsiderthecasewhenapartofmultilocusgenotypedataaremissingkeyword