SATS: A mutational signature analyzer of targeted sequenced tumors

Abstract

AbstractTumor mutational signatures are important in clinical decision-making and are typically analyzed using whole exome or genome sequencing (WES/WGS). However, targeted sequencing is more commonly used in clinical settings, posing challenges in mutational signature analysis due to sparse mutation data and non-overlapping targeted gene panels. We introduce SATS (Signature Analyzer for Targeted Sequencing), an analytical method that identifies mutational signatures in targeted sequenced tumors by analyzing tumor mutational burdens and accounting for different gene panels. We demonstrate through simulations and pseudo-targeted sequencing data (generated by down-sampling WES/WGS data) that SATS can accurately detect common mutational signatures with distinct profiles. Using SATS, we created a pan-cancer catalog of mutational signatures specifically tailored to targeted sequencing by analyzing 100,477 targeted sequenced tumors from the AACR Project GENIE. The catalog allows SATS to estimate signature activities even within a single sample, providing new opportunities for applying mutational signatures in clinical settings.