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 > DOI 10.1093bioinformaticsbtg1024 Predicting
侵权投诉

DOI 10.1093bioinformaticsbtg1024 Predicting

Vol. 19 Suppl. 1 2003, pages i183–i189

O.D.Kingetal.

DOI 10.1093bioinformaticsbtg1024 Predicting

DOI 10.1093bioinformaticsbtg1024 Predicting

DOI 10.1093bioinformaticsbtg1024 Predicting

beenusedpro tablyincollaborative lteringapplications(seee.g.Breeseetal.,1998).

Totesttheapproach,intheprocessofdoingthecross-validationwealsocompiledalistofthegene-phenotypepairs(i,j)∈G×TforwhichXj(i)=0butq(i,j)>0.5.Therewere542suchpairs,160ofwhichwerefor‘other’phenotypes(suchas‘otherDNAreplicationmutant’and‘othercytoskeletonmutants’)andwereremoved.Oftheremaining382pairs(i,j),welookedatthe100withthehighestscoresq(i,j)andmanuallyassessedtheplausibilityofgeneibeingassociatedwithphenotypejbylookingupgeneiintheYeastProteomeDatabase(Costanzoetal.,2001;doc.guandang.net),followingupinMEDLINEwhenwarranted.

Weusedafour-levelratingscheme.Thehighest-scoring50pairs(i,j)andtheirratingsaregiveninTable1—thefulllistof100ratings,alongwithothersupplementarydata,isavailableatdoc.guandang.net~king/pheno.html;belowwesummarizethenumberofgene-phenotypepairsthatreceivedeachrating:

Rating

(1)Nullmutanthasphenotype

(2)(Non-null)mutanthasphenotype(3)Nodecisiveevidence(4)Contradictoryevidence

Number

1726507

Theseresultsindicatethatoursuccessrateforthese100predictionswasatleast43%andatmost93%.

Fig.1.ROCcurvesforcross-validationofphenotypepredictions(1)usingbothMIPSannotationsandSGDGOannotationsaspredictors;(2)usingjustMIPSannotationsaspredictors;(3)usingjustSGDGOannotationsaspredictors;and(4)modelingallMIPSphenotypesandGOattributesasindependent.Inthetop gureweuseallgenesfortesting,andinthemiddle gureweuseonlythosegeneswithnoSGDGOannotationsoftypeIMP,TAS,NAS,ICorNR,toavoidcircularity.Thebottom gureisadetailofthemiddle gureatlowfalse-positiverates,withtheaxesrescaled.

Experimentalassessment

Wealsoexperimentallyassessed61ofthepredictedgene-phenotypeassociationsnotlistedinMIPS.Thesewerenotthe61predictionswiththehighestqscores,butwerethosepredictionswithq(i,j)>0.5andwithphenotypejamongelevenphenotypes(listedinTable2)withavailablehigh-throughputassayresults.Theseelevenphenotypesweretestedin4710yeastdeletionmutants(Winzeleretal.,1999;Giaeveretal.,2002)(ResearchGenetics),basedongrowthonsolidagarmediainvariousconditions.Furtherdetailsonthephenotypescreens,includingthequantitationandnormalizationofmutantgrowth,willbegiveninaseparatepublication(Dudley,Janse,andChurch;manuscriptinpreparation),butareavailableuponrequest.Webrie ydescribethemethodologybelow.

Allphenotypeswereexaminedinahomozygousdiploidbackground(BY4743).WiththeexceptionofYPG,whichcontained3%glycerolasthesolecarbonsource,allmediawerepreparedasYPD(Roseetal.,1990)containing2%glucosewiththeconcentrationofdrugsorchemicalsas

i186

上一页第4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