WCB ’ 11 Workshop on Constraint Based Methods for Bioinformatics September 12 th , 2011 Proceedings

A pathway is a group of genes which act together in a coordinated manner to control a cellular phenotype. The characterization of each set of genes forming a pathway and their organization inside of it represents an hard task in biology. Clonal outgrowth is a cellular phenotype resulting from a combination of acquired mutations affecting cellular genes. The insertion of proviruses in the genome of the host cell can mutate cellular genes and it might confer a selective advantage to the host cell resulting in clonal outgrowth. When the proviruses are experimentally injected into cells it takes the name of insertional mutagenesis. By this genetic approach we wanted to discover genome-wide the set of genes involved in the pathway controlling clonal outgrowth. We conducted insertional mutagenesis on mouse model of PML-RARalpha induced leukemia. Examining 48 leukemias we found a set of nearly 200 genes affected by the proviruses linked to clonal outgrowth. To infer the organization of these genes inside a pathway we took advantage by Mutual Information (MI) reverse-engineering approach [1] to quantify the extent to which the expression profiles of two genes are related to each other across a dataset of 20000 gene expression profiles. Out of the 200 genes, we discovered large and interconnected co-expressed modules consisting of genes strongly connected to each other. These coexpressed modules represent genes putatively sharing the same features and occupying the same position inside the pathway organization. The co-expressed modules constitute the basis to detail the fine organization and each biological role of the pathway components. We discuss the computational challenges and the future perspectives about insertional mutagenesis teqnique as an highly informative genetic approach to infer gene pathway linked to cellular proliferation.