文档介绍:*聚类与树Clustering*/54主要内容Microarrays(微阵列)HierarchicalClustering(层次聚类或系统聚类)K-MeansClustering(K-均值聚类)*/biningsimilardatapoints.*/paringthenewgenesequencestoknownDNAsequencesoftendoesnotgiveawaythefunctionofgeneFor40%ofsequencedgenes,paringtosequencesofotherknowngenesMicroarraysallowbiologiststoinfergenefunctionevenwhensequencesimilarityaloneisinsufficienttoinferfunction.*/54MicroarraysandExpressionAnalysisMicroarraysmeasuretheactivity(expressionlevel)ofthegenesundervaryingconditions/timepointsExpressionlevelisestimatedbymeasuringtheamountofmRNAforthatparticulargeneAgeneisactiveifitisbeingtranscribedMoremRNAusuallyindicatesmoregeneactivity*/54MicroarrayExperimentsProducecDNAfrommRNA(DNAismorestable)paremanysamplesatonceHybridizecDNAoverthemicroarrayScanthemicroarraywithaphosphor-illuminatinglaserIlluminationrevealstranscribedgenesScanmicroarraymultipletimesforthedifferentcolorphosphor’s*/54MicroarrayExperiments(con’t)phorscanbeaddedhereinsteadTheninsteadofstaining,laserilluminationcanbeused*/54UsingMicroarraysEachboxrepresentsonegene’sexpressionovertimeTrackthesampleoveraperiodoftimetoseegeneexpressionovertimeTracktwodifferentsamplesunderthesameconditionstoseethedifferenceingeneexpressions*/54UsingMicroarrays(cont’d)Green:expressedonlyfromcontrolRed:expressedonlyfromexperimentalcellYellow:equallyexpressedinbothsamplesBlack:NOTexpressedineithercontrolorexperimentalcells*/54MicroarrayDataMicroarraydataareusuallytransformedintoanintensitymatrix(below)Theintensitymatrixallowsbiologiststomakecorrelationsbetweendiferentgenes(eveniftheyaredissimilar)andtounderstandhowgenesfunctionsmightberelatedTime:(expressionlevel)ofgeneatmeasuredtime