Several recent studies have attempted the structure based prioritization of functional mutations in cancer. However, few have specifically explored the spectrum of somatic mu tations in protein pocket regions. In this study, we de veloped a protein structure based computational approach to explore the biochemical and structural roles of somatic mutations during tumorigenesis through the integration of large scale somatic mutation profiles onto protein pocket regions. The rationale of our computational approach is that if a gene has more somatic mutations in its protein pocket region, it is likely to be cancer related. To test this hypoth esis, we used three complementary methods cancer gene enrichment analysis we found that genes harboring somatic mutations in their protein pocket regions were significantly enriched with cancer genes.

functionally similar pair enrichment analysis in co expressed protein interaction networks genes harboring somatic mutations in their pocket regions tended to be highly co expressed in co expressed protein interaction networks. and anti cancer drug response gene enrichment analysis genes har boring somatic mutations in their protein pocket regions were more likely to be drug sensitive or drug resistant. Put together, somatic mutations located in protein pocket regions may be enriched with actionable mutations, and through their interactions drive tumorigenesis and alter anticancer drug treatment. To demonstrate the potential value of our approach, we identified four putative cancer genes, whose expres sion was associated with poor survival rates in melanoma, lung, or colon cancer patients.

Furthermore, in a case study using a protein pocket based approach rather than a traditional mutation versus wild type approach, we con cluded that the BAX gene was related to three anticancer drug sensitivities. There are two types of molecular mech anisms to explain mutations Entinostat in pocket residues are drug resistant or drug sensitive. A drug binds to a protein that directly involves the mutation in the pocket. For example, several independent studies found that the ac tionable mutations in the EGFR gene could activate EGFR by altering the ATP binding site, which ultimately leads to an enhancement of drug response to gefitinib. The pocket mutations affect protein function, which subsequently perturbs the network nodes in the drug targets signaling pathways, leading to drug sensitivity or resistance.

The second mechanism is in a ligand independent manner. Here, we did not find any direct evidence in that bcl 2 like protein 4 is a target pro tein involved in ligand protein binding with midostaurin, vinorelbine, or tipifarnib. Thus, BAX gene may perturb the network nodes in the signaling pathways, ultimately contributing to midostaurin, vinorelbine, and tipifarnib sensitivity.