pysigscore: gene signatures scoring across bulk and single-cell transcriptomics
Published in bioRxiv (under review), 2026
Gene signature scoring is a routine step in transcriptomics, but the choice of scoring method is rarely neutral: different methods can disagree substantially on the same data, and comparing them usually means stitching together several tools with incompatible interfaces.
pysigscore is a Python framework that brings 18 built-in scoring methods under a single API, together with a fully customisable scorer that lets users define and benchmark new scoring functions on equal footing. It works on both bulk and single-cell RNA-seq data.
Beyond computing scores, the package provides reliability analyses — p-value estimation and leave-one-out experiments — so that users can assess the significance of a score and understand how much individual genes contribute to it.
We validated pysigscore on the CCLE, TCGA and PBMC datasets, where it recovers the expected enrichment in liver, hypoxia, inflammatory and cell-cycle signatures. The package extends the functionality of the original R library sigscores and is distributed on PyPI.