Improving biomarker discovering for chemosensitivity prediction using an integrated approach

The discovery of genes that are responsible for cell responses to drugs, i.e., pharmacogenes, is a key task in pharmacogenomics. In this work we propose a new approach for pharmacogenes identification integrating: i) a filter and a wrapper feature selection technique, and ii) three sources of knowledge, namely transcriptional profiles, drugs activity, and path- way interactions. We show that the integration of multiple feature selection approaches as well as of several sources of knowledge increases the accuracy of the results.