蛋白质相互作用的生物信息学

蛋白质相互作用的生物信息学高友鹤中国医学科学院基础医学研究所蛋白质相互作用的生物信息学1.实验数据2.蛋白质相互作用数据库3.高通量实验数据的验证4.蛋白质相互作用网络5.计算预测蛋白质相互作用实验数据1.蛋白质相互作用的知识来源于实验。2.高通量地应用传统实验方法获取大量相互作用信息。3.高通量的数据需要验证。高通量实验方法CurrOpinStructBiol2003,13:377Yeasttwo-hybridassay•Benefits:–invivo.–Don’tneedpureproteins.–Don’tneedAb.•Drawbacks:–onlytwoproteinsaretestedatatime(nocooperativebinding);–ittakesplaceinthenucleus,somanyproteinsarenotintheirnativecompartment;anditpredictspossibleinteractions,butisunrelatedtothephysiologicalsetting.Massspectrometryofpurifiedcomplexes•Benefits:–severalmembersofacomplexcanbetagged,givinganinternalcheckforconsistency;–anditdetectsrealcomplexesinphysiologicalsettings.•Drawbacks:–itmightmisssomecomplexesthatarenotpresentunderthegivenconditions;–taggingmaydisturbcomplexformation;andlooselyassociatedcomponentsmaybewashedoffduringpurification.CorrelatedmRNAexpression•Benefits:–itisaninvivotechnique,albeitanindirectone;–andithasmuchbroadercoverageofcellularconditionsthanothermethods.•Drawbacks:–itisapowerfulmethodfordiscriminatingcellstatesordiseaseoutcomes,butisarelativelyinaccuratepredictorofdirectphysicalinteraction;–anditisverysensitivetoparameterchoicesandclusteringmethodsduringanalysis.Geneticinteractions(syntheticlethality).•Benefits:itisaninvivotechnique,albeitanindirectone;anditisamenabletounbiasedgenome-widescreens.•Drawbacks:notnecessarilyphysicalinteractions蛋白质相互作用的生物信息学1.实验数据2.蛋白质相互作用数据库3.高通量实验数据的验证4.蛋白质相互作用网络5.计算预测蛋白质相互作用蛋白质相互作用数据库CurrOpinStructBiol2003,13:377THEDIPDATABASE•DatabaseofInteractingProteins•TheDIPdatabasecatalogsexperimentallydeterminedinteractionsbetweenproteins.DIP相互作用的表达NucleicAcidsResearch,2000,28,289－291DIP数据库结构NucleicAcidsResearch,2000,28,289－291BIND：theBiomolecularInteractionNetworkDatabaseNucleicAcidsResearch,2001,29,242-245蛋白质相互作用的生物信息学1.实验数据2.蛋白质相互作用数据库3.高通量实验数据的验证4.蛋白质相互作用网络5.计算预测蛋白质相互作用高通量实验数据需要验证CurrOpinStructBiol2003,13:377与可信的数据相比CurrOpinStructBiol2003,13:377ExpressionProfileReliability•EPRIndexExpressionProfileReliabilityIndex(EPRIndex)evaluatesthequalityofalarge-scaleprotein-proteininteractiondatasetsbycomparingtheexpressionprofileoftheinteractingdatasetwiththatofthehigh-qualitysubsetoftheDIPdatabase.高通量数据互相比CurrOpinStructBiol2003,13:377ParalogousVerificationMethod•PVMScoreTheParalogousVerification(PVM)methodjudgesaninteractionprobableiftheputativelyinteractingpairhasparalogsthatalsointeract.DomainPairVerification•DPVScoreTheDomainPairVerification(DPV)methodjudgesaninteractionprobableifpotentialdomain-domaininteractionsbetweenthepairaredeemedprobable.CorrelationdistanceNatureBiotechnology2003,22,78蛋白质相互作用网络Nature2001,411,41-42相互作用网络的用途•Themosthighlyconnectedproteinsinthecellarethemostimportantforitssurvival.Nature2001,411,41-42蛋白质相互作用的生物信息学1.实验数据2.蛋白质相互作用数据库3.高通量实验数据的验证4.蛋白质相互作用网络5.计算预测蛋白质相互作用计算预测蛋白质相互作用CurrOpinStructBiol2003,13:377Docking•Need3DStructures•CAPRI:CriticalAssessmentofPredictedInteractions,acommunity-wideexperimentforassessingthepredictivepoweroftheseprocedures.ProteinFusion•Basedon:Somepairsofinteractingproteinsencodedinseparategenesinoneorganismarefusedtoproducesinglehomologousproteinsinotherorganism.•CompareE.Coliwithothergenomes:6,809putativeprotein-proteininteractionsMarcotteEMScience285,751(1999)•Compareyeastwithothers:45,502putativeinteractionsEnrightAJNature402,86(1999)GeneClustering•Basedon:Functionalcouplinggenesareinconservedgeneclustersindifferentgenomes.GeneClusteringOverbeekRPNAS96,2896(1999)OverbeekRPNAS96,2896(1999)PhylogeneticprofilePNAS(1999)96,4285-4288ACombinedExperimentalandComputationalStrategy•1)Screenrandompeptidelibrariesbyphagedisplaytodefinetheconsensussequencesforpreferredligandsthatbindtoeachpeptiderecognitionmodule.•2)Onthebasisoftheseconsensussequences,computationallyderiveaprotein-proteininteractionnetworkthatlinkseachpeptiderecognitionmoduletoproteinscontainingapreferredpeptideligand.Science2002295,321•3)Experimentallyderiveaprotein-proteininteractionnetworkbytestingeachpeptiderecognitionmoduleforassociationtoeachproteinoftheinferredproteomeintheyeasttwo-hybridsystem.•4)Determinetheintersectionofthepredictedandexperimentalnetworksandtestinvivothebiologicalrelevanceofkeyinteractionswithinthisset.ACombinedExperimentalandComputationalStrategyScience2002295,321高友鹤[e-mail]