Additional file 6: of GOexpress: an R/Bioconductor package for the identification and visualisation of robust gene ontology signatures through supervised learning of gene expression data

Comparison of the importance score from the random forest algorithm (i.e., the mean decrease in Gini index) to the F-ratio from a one-way ANOVA. Probesets with large importance score (i.e., good classifiers) generally show a high F-ratio, indicative of high variance among the means of each group compared to the variance within the samples. (PDF 1215 kb)