I currently have a data set composed of ~108 samples with RNAseq gene-counts and and ~20 clinical attributes (~75 diseased, ~33 healthy). The clinical data contains cell counts, and other clinical tests used to measure lung function. I have used WGCNA extensively, tried some basic regressions with Weka, performed differential expression, and hierarchical clustering. So far I have only verified what others have previously discovered. I was wondering if there is an R package or machine learning technique for combining both RNA-seq and clinical information to help identify novel disease sub-types?