文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > 自然科学 > 生物学 > EPIQ—efficient detection of SNP–SNP epistatic interactions for quantitative traits
侵权投诉

EPIQ—efficient detection of SNP–SNP epistatic interactions for quantitative traits

BIOINFORMATICSVol.30ISMB2014,pagesi19–i25

doi:10.1093/bioinformatics/btu261

EPIQ—efficientdetectionofSNP–SNPepistaticinteractionsforquantitativetraits

Ya’araArkin1,EliorRahmani1,MarcusE.Kleber2,ReijoLaaksonen3,4,WinfriedMarz€2,5,6andEranHalperin1,7,8,*

1TheBlavatnikSchoolofComputerScience,TelAvivUniversity,Tel-Aviv69978,Israel,2VthDepartmentofMedicine(Nephrology,Hypertensiology,Endocrinology,Diabetology,Rheumatology),MedicalFacultyofMannheim,UniversityofHeidelberg,MannheimD-68167,Germany,3ZoraBiosciencesOy,Espoo02150,Finland,4MedicalSchool,UniversityofTampere,Tampere33104,Finland,5ClinicalInstituteofMedicalandChemicalLaboratoryDiagnostics,Medical

UniversityofGraz,GrazA-8036,Austria,6SynlabAcademy,SynlabServicesGmbH,MannheimD-68165,Germany,7DepartmentofMolecularMicrobiologyandBiotechnology,GeorgeWiseFacultyofLifeScience,Tel-AvivUniversity,Tel-Aviv69978,Israeland8InternationalComputerScienceInstitute,Berkeley,CA94704,USA

ABSTRACTphenomenaoftenreferredtoasthe‘missingheritability’Motivation:Gene–geneinteractionsareofpotentialbiologicaland(Maher,2008).Oneplausibleexplanationsuggestedforthismedicalinterest,astheycanshedlightonboththeinheritancemech-problemistheexistenceofanepistaticeffect,wheretwooranismofatraitandontheunderlyingbiologicalmechanisms.morelocihaveasynergeticinfluenceonthephenotype,alsoEvidenceofepistaticinteractionshasbeenreportedinbothhumansreferredtoasgene–geneinteractions(Maher,2008).Thediscov-andotherorganisms.Unlikesingle-locusgenome-wideassociationeryofinteractingSNPlocihasanadditionalbenefit,asitmaystudies(GWAS),whichprovedefficientindetectingnumerousgeneticshedlightontheunderlyingbiologicalmechanismorinvolvedlocirelatedwithvarioustraits,interaction-basedGWAShavesofarpathways.

producedveryfewreproduciblediscoveries.SuchstudiesintroduceaDespiteevidencesofgene–geneinteractionsreportedinbothgreatcomputationalandstatisticalburdenbynecessitatingalargehumanandotherorganisms(Evansetal.,2006),veryfewrepro-numberofhypothesestobetestedincludingallpairsofsinglenucleo-duciblediscoverieswerereportedbyGWAS(Liuetal.,2011;tidepolymorphisms(SNPs).Thus,manysoftwaretoolshavebeenPrabhuandPe’er,2012forexample).Theamountofdatapro-developedforinteraction-basedcase–controlstudies,someleadingducedinasinglestudyisapossiblecause:whensearchingfortoreliablediscoveries.Forquantitativedata,ontheotherhand,onlyagroupsofkSNPswithanepistaticeffect,thenumberofpossiblehandfuloftoolsexist,andthecomputationalburdenisstillsubstantial.k-sizedgroupsisÂ(mk),wheremisthenumberofSNPloci.Results:WepresentanefficientalgorithmfordetectingepistasisinWithcurrentGWAStypicallyincludinghundredsofthousandsquantitativeGWAS,achievingasubstantialruntimespeedupbyavoid-ofSNPs,thisimpliesbothacomputationalandstatisticalburdeningtheneedtoexhaustivelytestallSNPpairsusingmetricembeddingevenforgroupssizesassmallask=2:thenumerousteststakesandrandomprojections.Unlikepreviousmetricembeddingmethodsdaysandevenweekstocomputeandrequireasubstantialcor-forcase–controlstudies,weintroduceanewembedding,whereeachrectionformultiplehypothesis,leadinginsomecasestoalossofSNPismappedtotwoEuclideanspaces.Weimplementedourpower(Evansetal.,2006).Onecommonapproachisthereduc-methodinatoolnamedEPIQ(EPIstasisdetectionforQuantitativetionofthesearchspace,usuallybyfilteringcandidatelocipairs:GWAS),andweshowbysimulationsthatEPIQrequireshoursofpro-Marchinietal.(2005)suggestedselectingasubsetofSNPswithacessingtimewhereothermethodsrequiredaysandsometimesmoderatemarginaleffectandtestingforinteractioninpairsweeks.ApplyingourmethodtoadatasetfromtheLudwigshafenwhereatleastonelocusisincludedinthesubset.Reductionofriskandcardiovascularhealthstudy,wediscoveredapairofSNPsthesearchspacecanalsobedonebymanipulatingcontingencywithanear-significantinteraction(P=2.2Â10À13),inonly1.5hon10tables(Wanetal.,2010;Zhangetal.,2010)orsearchingforaprocessors.linkagedisequilibrium(LD)contrastbetweencasesandcontrolsAvailability:doc.guandang.net(Brinzaetal.,2010;PrabhuandPe’er,2012).Amorestraight-Contact:heran@post.tau.ac.ilforwardapproachisincreasingthecomputationalpower,either

bymulti-threadedimplementationsorbyutilizingspecialhard-

1INTRODUCTIONware(Huetal.,2010;Yungetal.,2011).Binaryoperationsare

usedinsomecasestospeedupperformance(PrabhuandPe’er,

Genome-wideassociationstudies(GWAS)havesofardetected2012;Wanetal.,2010).

thousandsofsinglenucleotidepolymorphism(SNP)locithatareAllofthesetools,andmanyothers,aredesignedforcase–associatedwithvarioustraits(Hindorffetal.,2009).controlstudies;whereasforthequantitativecase,wheretheUnfortunately,formostcomplextraitsthediscoveredSNPstestedphenotypesarephysiologicalmeasurementsofsomeexplainonlyasmallfractionoftheestimatedheritability,asort,theselectionofavailablesoftwareislimited.Sincethe

phenotypetestedisnotdichotomous,testingforquantitative

associationscanbemorechallengingcomparedwithcase–con-

trolstudies,asmethodsutilizingcontingencytables,LD-contrast

*Towhomcorrespondenceshouldbeaddressed.orbinaryoperationsareusuallyinapplicable.MethodstailoredßTheAuthor2014.PublishedbyOxfordUniversityPress.

ThisisanOpenAccessarticledistributedunderthetermsoftheCreativeCommonsAttributionNon-CommercialLicense(doc.guandang.net),whichpermitsnon-commercialre-use,distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.Forcommercialre-use,pleasecontactjournals.permissions@oup.comDownloaded from doc.guandang.net at XidianUniversity on December 1, 2014

第1页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email