Genetic instability is one of the hallmarks of cancer. Neoplastic cells accumulate somatic mutations in their genomes, resulting in aberrant homeostasis, cancer cell survival, and proliferation. Different genetic instability processes result in distinct non-random patterns of DNA mutations, also known as mutational signatures. The interest in the identification of mutational signatures and the corresponding genetic instability processes is rapidly growing because these signatures are footprints of the molecular aberrations occurring in tumors, and hence may be prognostic of clinical outcomes and support customized anti-cancer treatments in the future. We developed ‘mutSignatures’, an integrated R-based computational framework aimed at deciphering DNA mutational signatures. Our software provides advanced functions for importing DNA variants, computing tri-nucleotide or non-standard mutation types, and extracting mutational signatures via non-negative matrix factorization. Additionally, our framework supports deconvolution of catalogs of DNA variants against known mutational signatures. We used ‘mutSignatures’ to analyze somatic mutations from smoking-related cancer datasets, revealing mutational signatures that were consistent with those reported before in independent investigations. Moreover, our analyses showed that selected mutational signatures correlated with specific clinical and molecular features. Specifically, we analyzed mutations from a lung adenocarcinoma dataset, and identified two signatures, namely ‘luad_B’ and ‘luad_C’, with opposite trends: signature ‘luad_B’ was increased in tumors from smokers and correlated with high mutation burden; conversely, signature ‘luad_C’ was enriched in tumors from life-long non-smokers, and correlated with low mutation burden. These signatures may represent the product of mutually-exclusive processes that are responsible of initiating lung neoplastic transformation in smoking and non-smoking patients. In conclusion, we developed ‘mutSignatures’, a powerful open-source framework for extraction and analysis of mutational signatures. Our software enables the study of the molecular determinants of cancer, supports the analysis of signature-related clinical characteristics, and could facilitate gathering insights into cancer biology and treatment.