As part of the publication in Science of our latest findings in the field of mutational signatures, we have released an updated version of our R package, signature.tools.lib, which includes our new signature fit algorithm FitMS, alongside multiple other improvements (version is v2.1.2).

We have a new basic signature fit function called Fit, which replaces the v1.0 fit functions signatureFit and signatureFit_withBootstrap. We have also improved the plotting of signature fit results, whether from Fit or FitMS, which in the new version can be done using the plotFitResults function.

On top of the Fit and FitMS functions we have also written the signatureFit_pipeline function, which is a unified interface for signature fitting, and which simplifies generating catalogues and selecting signatures for fitting. A command line script, signatureFit, is also available as a wrapper of the signatureFit_pipeline R function.

This means that running a signature analysis using signature.tools.lib and FitMS can be as simple as typing:

signatureFit -O Breast -x snvvcf.tsv

The option -O indicates the tissue-type of origin of the samples, thus indicating which signatures FitMS should use, and the file snvvcf.tsv should contain a list of VCF file names containing single nucleotide variants, one VCF file for each sample to be analysed.

The latest version of signature.tools.lib, along with documentation and examples, can be found here.