Mutational signatures are characteristic combinations of mutation types arising from specific mutagenesis processes such as DNA replication infidelity, exogenous or endogenous exposures, defective DNA repair or DNA and RNA enzymatic editing. Analysis of mutational signatures is becoming routine in cancer genomics, providing a great opportunity for biomarker discovery, tumor diagnostics and treatment guidance.However, existing software packages (or mutational signature visualization software) have several drawbacks. Firstly, they are not designed for large-scale analysis of multiple datasets at the same time. Further, existing packages show different performance and accuracy in extracting known mutational signatures and there are no gold standards to benchmark different methods/algorithms. Moreover, existing packages can only identify signatures, but not carry out integrative analyses of signatures together with other tumor genetic or molecular features, or association analyses with clinical data. Most importantly, these packages require computational and bioinformatic expertise, and are thus not accessible to the broader research community without such expertise. While web-accessible analysis portals such as cBioPortal have greatly facilitated integrative analysis of TCGA and other tumor datasets, no such accessible and user-friendly tool is available for integrative analysis of mutational signatures.We developed an integrative mutational signature portal (mSigPortal) to explore data from published genomic studies and/or analyze user's data. mSigPortal allows the following functions : 1) To interactively and comprehensively visualize spectra of mutation types, copy number alterations, structural variants and other genomic features at both sample and study level, as well as to conduct profile comparisons, principal component analysis and motif enrichment analyses. 2) To interactively explore both published signatures and processes from different genomic studies (e.g., The Cancer Genome Atlas (TCGA), Pan-Cancer Analysis of Whole Genomes (PCAWG), and the Sherlock-Lung study) and to decompose the mutational patterns and other genomic features into accurate mutational signatures. 3) To comprehensively evaluate the accuracy of mutational signatures based on multiple statistical factors and suggest the best solution based on these factors and different algorithms (e.g., SigProfiler, deconstructsig, bootstrapping method). 4) To systematically analyze and visualize the association between specific exposure mutational signatures and genomic or epigenomic features or clinical data from different genomic studies. mSigPortal provides state-of-the-art methods and an easily accessible platform to explore, visualize and analyze mutational signatures, which will greatly facilitate studies of mutagenetic processes involved in tumorigenesis.

Citation Format: Tongwu Zhang, Jian Sang, Phyllip Cho, Kevin Jiang, Maria Teresa Landi. Integrative mutational signature portal (mSigPortal) for cancer genomic study [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 211.