SigProfilerWeb: a user-friendly web application for accurately determining the activities of mutational signatures

Marcos Díaz-Gay1, S. M. Ashiqul Islam2, Maria Vila-Casadesús3, Sebastià Franch-Expósito1, Juan José Lozano4, Ludmil B. Alexandrov5, Sergi Castellví-Bel1

1Gastroenterology Department, Hospital Clínic, Institut d’Investigacions Biomèdiques August Pi i Sunyer, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas, Barcelona, Spain. 2Department of Cellular and Molecular Medicine, Department of Bioengineering, Moores Cancer Center, University of California, San Diego, La Jolla, USA. 3Gene Regulation, Stem Cells and Cancer Program, Centre for Genomic Regulation, The Barcelona Institute of Science and Technology, Barcelona, Spain. 4Bioinformatics Platform, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas, Barcelona, Spain. 5Department of Cellular and Molecular Medicine, Department of Bioengineering, Moores Cancer Center, University of California, San Diego, Barcelona, Spain

Abstract

OBJECTIVES

Endogenous and exogenous mutational processes are moulding the genomes of somatic cells and imprinting mutational patterns, also known as mutational signatures. Analysis of mutational signatures has proven as a valuable tool in cancer genomics with potential applications ranging from identifying different germline deficiencies (e.g., NTHL1 deficiencyto revealing clinically relevant biomarkers (e.g., susceptibility to PARP inhibitors). The purpose of this work is to facilitate the implementation of this methodology in a single sample environment, suitable for clinical practice.

 

METHODS

Different sets of mutational signatures were deciphered for distinct types of somatic mutations, including single base substitutions (SBS), doublet base substitutions (DBS) and small insertions and deletions (ID). The COSMIC database was recently updated to reflect the analysis of 23,829 cancer genomes used by the non-negative matrix factorisation-based software SigProfiler. A total of 77 biologically relevant consensus mutational signatures (49 SBS, 11 DBS, and 17 ID signatures) were reported.

 

RESULTS

In this work we present SigProfilerWeb, a web application based on the Shiny R framework to implement the single sample algorithm of SigProfiler in the form of a user-friendly interactive tool. A complete characterisation of the somatic profile is provided, including calculation of tumor mutational burden, mutational profile plotting for the different mutation classes, transcriptional strand bias analysis, signature refitting and sample classification using clustering and principal component analysis.

 

CONCLUSIONS

SigProfilerWeb is a web application that provides advanced capabilities for analysis of mutational signatures in both research and clinical settings. SigProfilerWeb will be released soon as a freely accessible website.

Top