̽����ѡ

Cancer is a unique disease in which cells acquire positive traits for survival, mirroring species evolution. Classical methods to identify driver mutations aim to detect the signatures of Darwinian positive selection, which appear as 3D-clusters of dozens of mutations in localized regions of protein structures. Such methods cannot explain the reasons for mutational clustering, which requires understanding how the protein structure is linked to function: the so-called conformational changes. Here we propose to apply an innovative methodology that pushes genomic data towards conformational mutation profiling using simulations, structural techniques, and molecular biology. We have successfully applied this approach to the oncogene EGFR, revealing how heterogeneous missense and deletion mutations evolutionarily converge in glioblastoma to a similar structural-functional effect, and thus respond to the same drugs. Our findings shed light on how apparently diverse mutations can be biochemically equivalent, reducing the complexity of tumor heterogeneity to fewer evolutionarily selected traits. Our preliminary results indicate evolutionarily convergent structural traits appear in many other unexpected proteins such as sugar transporters, channels and pumps. This project aims to explore in the extent and role of molecular convergence in cancer in two phases -computational and experimental- with the goal to understand the biological effect of mutations and rationalize cancer therapy.