An Integrated Approach for Identifying Wrongly Labelled Samples When Performing Classification in Microarray Data 英文参考文献.docVIP

AnIntegratedApproachforIdentifyingWrongly LabelledSamplesWhenPerformingClassificationin MicroarrayData YukYeeLeung1,2,3*,ChunQiChang1,YeungSamHung1 1DepartmentofElectricalandElectronicEngineering,TheUniversityofHongKong,HongKongSpecialAdministrativeRegion,China,2DepartmentofPathologyand Laboratory Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, United States of America, 3Penn Center for Bioinformatics, University of Pennsylvania, Philadelphia,Pennsylvania,UnitedStatesofAmerica Abstract Background:Usinghybridapproachforgeneselectionandclassificationiscommonasresultsobtainedaregenerallybetter thanperformingthetwotasksindependently.Yet,forsomemicroarraydatasets,bothclassificationaccuracyandstabilityof genesetsobtainedstillhaveroomsforimprovement.Thismaybeduetothepresenceofsampleswithwrongclasslabels (i.e. outliers). Outlier detection algorithms proposed so far are either not suitable for microarray data, or only solve the outlierdetectionproblemontheirown. Results:WetackletheoutlierdetectionproblembasedonapreviouslyproposedMultiple-Filter-Multiple-Wrapper(MFMW) model,whichwasdemonstratedtoyieldpromisingresultswhencomparedtootherhybridapproaches(LeungandHung, 2010). To incorporate outlier detection and overcome limitations of the existing MFMW model, three new features are introducedinourproposedMFMW-outlierapproach:1)anunbiasedexternalLeave-One-OutCross-Validationframeworkis developed to replace internal cross-validation in the previous MFMW model; 2) wrongly labeled samples are identified withintheMFMW-outliermodel;and3)astablesetofgenesisselectedusinganL1-normSVMthatremovesanyredundant genes present. Six binary-class microarray datasets were tested. Comparing with outlier detection studies on the same datasets, MFMW-outlier could detect all the outliers found in the original paper (for which the data was provided for analysis),andthegenesselectedafteroutlierremovalwereproventohavebiologicalrelevance.WealsocomparedMFMW- outlierwithPRAPIV(Zhangetal.