Supervised Methods with Genomic Data: a Review and Cautionary View

摘要: We review well accepted methods to address questions about differential expression of genes and class prediction from gene data. highlight some new topics that deserve more attention: testing specific groups genes, intra-group heterogeneity prediction, interaction in predictors, visualisation, difficulties the biological interpretation predictor molecular signatures, use ROC[Receiver Operating Characteristic curve]-based statistics for evaluating predictors expression. end with a serious problems can limit potential these methods; we focus specially on inadequate assessment performance (due estimation error rates few “easy” data sets) failure recognise observational studies include needed covariates. A final comment is made need freely available source code.