Abstract

Gene expression profiles are unveiling a wealth of new potential drug targets for a wide range of diseases, offering new opportunities for drug discoveries. The emerging challenge, however, is the effective selection of the myriad of targets to identify those with the most therapeutic utility. Numerical clustering has became a commonly used method to investigate and interpret gene expression data sets but it is often inadequate to infer the genes and proteins role and point to candidate genes for drug development. This review illustrates how clustering methods based on semantic characteristics, such as gene ontologies, could be used to extract more knowledge from genomic data and improve drug target and discovery processes.

Keywords: clustering, drug, genomics, microarray, semantic