Machine learning for molecular systems biology

Abstract

The development of high-throughput techniques in biology is transforming biology in a data-rich discipline. We will consider in this project integrated biological networks: these networks deal with all the gene interactions mediated by metabolism, regulation and protein-protein interactions. We will deploy machine learning tools that will use topological data from these graphs, expression data, genomic organization and cellular localization to extract relevant biological information such as detection of drug target genes, morbid genes for humans or essential genes for bacteria. We will also investigate the community structure of these networks and we will detect topological motifs aiming the characterization of the large scale topological organization of gene interactions on biological cells. (AU)