Cancer genomes accumulate many somatic mutations resulting from carcinogenic exposures, cancer-related aberrations of DNA maintenance machinery, and normal stochastic events. These processes often lead to distinctive patterns of mutations, called mutational signatures. However interpreting mutation patterns captured by such signatures is often challenging as specific patterns observed in cancer genomes often emerge as the end effect of interactions between DNA damage, DNA repair, and other molecular processes. Untangling these contributions and identifying interactions between mutagenic and other cellular processes remains difficult. In this talk, I will describe computational methods to infer the relations between mutational signatures and cellular and environmental processes.