articles
VirtualScreeningofSelectiveMultitargetKinaseInhibitorsbyCombinatorialSupportVectorMachines
X.H.Ma, , R.Wang, C.Y.Tan,§Y.Y.Jiang,§T.Lu,|H.B.Rao,|X.Y.Li,|
M.L.Go,⊥B.C.Low, andY.Z.Chen*, ,§,|
BioinformaticsandDrugDesignGroup,DepartmentofPharmacy,CentreforComputationalScienceandEngineering,NationalUniVersityofSingapore,BlkS16,LeVel8,3ScienceDriVe2,Singapore117543,TheKeyLaboratoryofChemicalBiology,GuangdongProVince,TheGraduateSchoolatShenzhen,TsinghuaUniVersity,Shenzhen,Guangdong,P.R.China,CollegeofChemicalEngineeringandStateKeyLaboratoryofBiotherapy,SichuanUniVersity,Chengdu610065,P.R.China,DepartmentofPharmacy,NationalUniVersityofSingapore,ScienceDriVe4,Singapore117543,andDepartmentofBiologicalScience,NationalUniVersity
ofSingapore,BlkS2,LeVel5,ScienceDriVe4,Singapore117543
ReceivedAugust31,2009;RevisedManuscriptReceivedAugust7,2010;AcceptedAugust16,2010
Abstract:Multitargetagentshavebeenincreasinglyexploredforenhancingef cacyandreducing
countertargetactivitiesandtoxicities.Ef cientvirtualscreening(VS)toolsforsearchingselectivemultitargetagentsaredesired.Combinatorialsupportvectormachines(C-SVM)weretestedasVStoolsforsearchingdual-inhibitorsof11combinationsof9anticancerkinasetargets(EGFR,VEGFR,PDGFR,Src,FGFR,Lck,CDK1,CDK2,GSK3).C-SVMtrainedon233-1,316non-dual-inhibitorscorrectlyidenti ed26.8%-57.3%(majority>36%)ofthe56-230intra-kinase-groupdual-inhibitors(equivalenttothe50-70%yieldsoftwoindependentindividualtargetVStools),and12.2%ofthe41inter-kinase-groupdual-inhibitors.C-SVMwerefairlyselectiveinmisidentifyingasdual-inhibitors3.7%-48.1%(majority<20%)ofthe233-1,316non-dual-inhibitorsofthesamekinasepairsand0.98%-4.77%ofthe3,971-5,180inhibitorsofotherkinases.C-SVMproducedlowfalse-hitratesinmisidentifyingasdual-inhibitors1,746-4,817(0.013%-0.036%)ofthe13.56MPubChemcompounds,12-175(0.007%-0.104%)ofthe168KMDDRcompounds,and0-84(0.0%-2.9%)ofthe19,495-38,483MDDRcompoundssimilartotheknowndual-inhibitors.C-SVMwascomparedtootherVSmethodsSur ex-Dock,DOCKBlaster,kNNandPNNagainstthesamesetsofkinaseinhibitorsandthefullsetorsubsetofthe1.02MZincclean-leadsdataset.C-SVMproducedcomparabledual-inhibitoryields,slightlybetterfalse-hitratesforkinaseinhibitors,andsigni cantlylowerfalse-hitratesfortheZincclean-leadsdataset.CombinatorialSVMshowedpromisingpotentialforsearchingselectivemultitargetagentsagainstintra-kinase-groupkinaseswithoutexplicitknowledgeofmultitargetagents.
Keywords:Anticancer;multitarget;high-throughputscreening;computeraideddrugdesign;kinaseinhibitor;supportvectormachines;virtualscreening
Introduction
Alargepercentageofdrugsindevelopment,whicharetypicallydirectedatanindividualtarget,frequentlyshow
*Correspondingauthor.Mailingaddress:NationalUniversityofSingapore,DepartmentofPharmacy,BlkS16,Level8,ScienceDrive2,Singapore117543.Tel:65-6874-6877.Fax:65-6774-6756.E-mail:phacyz@nus.edu.sg.
BioinformaticsandDrugDesignGroup,DepartmentofPhar-macy,CentreforComputationalScienceandEngineering,NationalUniversityofSingapore.
DepartmentofBiologicalScience,NationalUniversityofSingapore.
10.1021/mp100179t 2010AmericanChemicalSociety
reducedef caciesandundesiredsafetyandresistancepro lesduetonetworkrobustness,1redundancy,2crosstalk,3com-§|
TsinghuaUniversity.SichuanUniversity.⊥
DepartmentofPharmacy,NationalUniversityofSingapore.
(1)Smalley,K.S.;Haass,N.K.;Brafford,P.A.;Lioni,M.;Flaherty,
K.T.;Herlyn,M.Multiplesignalingpathwaysmustbetargetedtoovercomedrugresistanceincelllinesderivedfrommelanomametastases.Mol.CancerTher.2006,5(5),1136–44.
(2)Pilpel,Y.;Sudarsanam,P.;Church,G.M.Identifyingregulatory
networksbycombinatorialanalysisofpromoterelements.Nat.Genet.2001,29(2),153–9.VOL.7,NO.5,1545–1560MOLECULARPHARMACEUTICS1545
PublishedonWeb08/16/2010
Synthesis and Characterization of Pyrazoleanthrone Derivatives as Aurora A Kinase Inhibitors_生产/经营管理_经管营销_专业资料。Chem.Res.Chin.Univ.2013.29(6)...
Synthesis and Characterization of Pyrazoleanthrone Derivatives as Aurora A Kinase Inhibitors_生产/经营管理_经管营销_专业资料。Chem.Res.Chin.Univ.2013.29(6)...
热门文档