Read more
Over the past decade, microarray gene expression datahas been widely used to identify cancer genesignatures that could complement conventionalhistopathologic evaluation to increase the accuracyof cancer diagnosis and prognosis. However, due tothe limited sample size of individual studies, thereis often only a small overlap between signatures forspecific cancers obtained in different studies. Withthe rapid accumulation of microarray data, it is ofgreat interest to understand how to combinemicroarray data across studies of similar cancers inorder to increase sample size, which could lead tothe identification of more reliable gene signaturesof specific cancers. This book provides a novelstatistical method, based on the top-scoring pairsclassifier family, for inter-study microarray dataintegration. By incorporating this new method intodifferent statistical models, we can identify robustgene signatures for cancer diagnosis and prognosisfrom integrated microarray data. This book isprimarily aimed at cancer researchers, especially inthe emerging field of cancer bioinformatics, butshould also be valuable to statisticians and medicalresearchers entering the field.
About the author
PhD in Electrical Engineering with concentration on
Bioinformatics at Johns Hopkins University. Statistician in
biotechnology and pharmaceutical industry.
