Abstract 5722: The mutational signatures of 100,477 targeted sequenced tumors

Abstract Mutational signatures, the footprints of somatic mutations, are associated with the causes of cancer and have been well-studied for tumors with whole exome or genome sequencing. However, due to the low number of detected mutations, mutational signatures were insufficiently explored for many tumors sequenced by targeted panels in clinics. It impedes the clinical application of mutational signatures. Here, we present a new method, SALMON (Signature Analyzer for Low Mutation cOuNts), to identify mutational signatures in targeted sequenced tumors based on tumor mutational burden. SALMON adjusts for panel size differences and uses a large number of targeted sequenced tumors rather than a large number of mutations per tumor (as with whole exome or genome sequencing) to overcome the challenges of mutational signature analysis for targeted sequencing. Extensive simulations and pseudo-targeted sequenced data show that SALMON can accurately detect spiky or common signatures. We applied SALMON to investigate the pan-cancer patterns of mutational signatures for 100,477 targeted sequenced tumors in AACR Project GENIE, including 14,428 lung and 11,389 breast tumors. We detected well-established signatures in tumor types that have not previously been associated with these signatures, such as the smoking signature in ovarian tumors. Interestingly, analysis of thousands of tumors per cancer type from diverse populations revealed gender discrepancies, self-described race differences, subtype heterogeneity, and metastatic enrichment of mutational signatures. For instance, most sex-biased signatures are more frequently present in males for non-gender-specific cancers. Thiopurine treatment-induced signature in glioma is enriched in Black patients. And endometrioid ovarian or uterine cancers have a higher prevalence of polymerase epsilon (POLE) deficiency-related signatures than non-endometrioid ovarian or uterine cancers, respectively. Our study demonstrates the feasibility and utility of mutational signature analysis for targeted sequenced tumors, enabling precision applications of mutational signatures in the clinical setting. Citation Format: Donghyuk Lee, Min Hua, Difei Wang, Lei Song, Tongwu Zhang, Kai Yu, Xiaohong R. Yang, Jianxin Shi, Stephen J. Chanock, Maria Teresa Landi, Bin Zhu. The mutational signatures of 100,477 targeted sequenced tumors. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5722.