BIOINFORMATICSAPPLICATIONSNOTEBIOINFORMATICSAPPLICATIONSNOTEVol.21no.112005,pages2789–2790doi:10.1093/bioinformatics/bti394GeneexpressionMADE4:anRpackageformultivariateanalysisofgeneexpressiondataAedínC.Culhane1,?,JeanThioulouse2,GuyPerrière2andDesmondG.Higgins11Bioinformatics,ConwayInstitute,UniversityCollegeDublin,Dublin4,Irelandand2LaboratoiredeBiométrieetBiologieévolutive,UniversitéClaudeBernard,Lyon1,43,bd.du11Novembre1918,69622VilleurbanneCedex,FranceReceivedonFebruary9,2005;revisedandacceptedonMarch16,2005AdvanceAccesspublicationMarch29,2005ABSTRACTSummary:MADE4,microarrayade4,isasoftwarepackagethatfacilitatesmultivariateanalysisofmicroarraygene-expressiondata.MADE4acceptsawidevarietyofgene-expressiondataformats.MADE4takesadvantageoftheextensivemultivariatestatisticalandgraphicalfunctionsintheRpackageade4,extendingtheseforapplicationtomicroarraydata.Inaddition,MADE4providesnewgraphicalandvisualizationtoolsthataidininterpretationofmultivar---本文来源于网络，仅供参考，勿照抄，如有侵权请联系删除---iateanalysisofmicroarraydata.Availability:TheRpackageMADE4isavailablefromBiocon-ductorbioinf.vcd.ie/softwareandfromBioconductorwws://wwsodocs/doc/6415481049Contact:aedin.culhane@ucd.ieSupplementaryinformation:MADE4iswelldocumented.Therearetutorials,intheformofvignettes,whichdescribetypicalanalyses.Inaddition,theMADE4manualprovidesdescriptionsandexamplesforeachfunction.1INTRODUCTIONTheaiminwritingmicroarrayade4(MADE4)wastoprovideasimple-to-usetoolformultivariateanalysisofmicroarraydata.Multivariateapproacheshavebeenappliedverysuccessfullyintheanalysisofmicroarraydata.Principalcomponentanalysis(PCA)hasbeenshowntobeusefulinexploratoryanalysisoflineartrendsindata(Raychaudhurietal.,2000;CrescenziandGiuliani,2001).Fellenbergetal.(2001)describedtheapplicationofcorrespondenceanalysistostudytheassociationbetweenmicroarraysamplesandgenesinareduceddimensionalspace.Agroupordinationapproachwasappliedsuccessfullytoclassi?cationandclasspredictionofmicroarraysamples(Culhaneetal.,2002).Morerecently,Culhaneetal.(2003)employedatwo-tablecouplingmethod(coinertiaana-lysis,CIA)toexaminecovariantgene----本文来源于网络，仅供参考，勿照抄，如有侵权请联系删除---expressionpatternsbetweenmicroarraydatasetsfromdifferentplatforms.AlthoughPCAisavailableinseveralRpackages,includingstatsandamap,theRpackageade4containsmanyadditionalmultivariatestatisticalmethodsincludingmethodsforanalysisofone-datamatrix,couplingoftwo-datamatricesormulti-tableanalysis,cran.univ-lyon1.fr/doc/Rnews/4Rnews_2004-1.pdf(Thioulouseetal.,11017;Chesseletal.,2004).Theselattermeth-odsforintegratingmultipledatasetsmakethisparticularpackageTowhomcorrespondenceshouldbeaddressed.veryattractiveforanalysisofmicroarraydata.MADE4isdevelopedasanextensiontoade4tofacilitateinputandanalysisofmicroarraydata.Inordertoprovidethisfunctionality,MADE4isintegratedwithBioconductor(Gentlemanetal.,2004),probablythemostpopularmicroarrayanalysissoftware,whichcontainsnumerouspackagesforpreprocessing,normalization,gene?lteringandanalysisofmicroarraydata.2DATAINPUTMADE4acceptsawidevarietyofgene-expressiondatainputformats,includingBioconductorAffyBatch,exprSet,marrayRaw,andstandardRmatrixformats(data.frameormatrix).MADE4willautomaticallyrecognizethesedataformats,andnoadditionaldataprocessingisrequired.3MULTIVARIATEANALYSI...