Scientific sessions, CRG Group Leader Seminars
Genomic and Epigenomic Variation in Disease Group, Bioinformatics and Genomics Programme, CRG
Stephan Ossowski obtained a PhD in Computational Biology from the Max-Planck Institute in Germany in 2010 for his work on the computational analysis of next generation sequencing (NGS) data, genome re-sequencing and population genetics. In 2008 he published the first whole-genome analysis of a plant genome using NGS technology and has since worked on developing novel computational and statistical methods for NGS analysis. In 2010 he started the Genomic and Epigenomic Variation in Disease group at the Centre for Genomic Regulation in Barcelona. The lab is developing experimental and computational methods for medical genomics and epigenomics, with a strong focus on NGS applications in clinical diagnostics and precision medicine.
In this talk I will discuss novel computational methods we developed to identify and utilize signatures of positive and negative (purifying) selection in tumor evolution. Tumors evolve over time and accumulate somatic mutations. Acquisition of alterations conferring a selective growth advantage to neoplastic cells over surrounding normal or tumor cells leads to a rapid increase of the mutated sub-clone within the tumor, resulting in tumor heterogeneity and leading to treatment resistance and relapse. We have developed a statistical model to exploit signatures of tumor clonal evolution, e.g. the cancer cell fraction of somatic mutations, for identification of tumor driver genes. Applying our model to NGS data from 22 cancers studied as part of ICGC/TCGA we demonstrate the high accuracy of the novel approach. We further show that cancer drivers affecting a large fraction of patients in one cancer type can also play a role in a small fraction of patients in other cancers, but are often missed due to low recurrence.
Finally I will introduce our latest research on the identification of genes that are essential for the developing tumor and are thus depleted for damaging mutations. We show that mutations in these genes under purifying selection are detrimental for the tumor and increase the survival rate among affected patients.