Cancers are induced by the accumulation of genetic alterations within a cell,including inherited genetic mutations,chromosome translocations, andcopynumber alterations. There are many types of cancerdrugs,anti-cancer drugsensitivity study is of great significance. Considering the regulation functionbetween geneswe propose a network lfow-based method to predict anticancerdrug sensitivity by making use of the topological structure of pathways.According to our model, mutations and copy number alternations of cancerdriver genes of some cancer-related pathways are assumed to have certaincontributions to the pathway activity. Anti-cancer drugs would reduce a certainpercentage of pathway activity lfowing through its target genes. Sensitivity of adrug to a given sample is measured by the difference of pathway activities beforeand after drug treatment. Based on these assumptions, we come up with anoptimization model to ift all parameters based on known samples. Bycross-validation test, our model achieved good performance for drugs targetingspecifically to MAPK pathway. Our method is also potentially applicable topredict combination effects of different drugs. |