SelectSim infers evolutionary dependencies — co-mutations and mutual exclusivities — between functional alterations across cancer genomes. It estimates the expected co-mutation frequency for each gene pair from individual mutation frequencies and per-sample tumor mutation burden (TMB), then evaluates significance against a simulation-based null model.
This package accompanies the manuscript:
Iyer A, Petrovic M, Sesia D, Nanni L, Mina M, Ciriello G (2026). Evolving patterns of co-mutations from tumor initiation to metastatic progression. Nature Genetics. DOI: 10.1038/s41588-026-02661-4
cd SelectSim_py
uv sync # recommended; or: pip install -e .Optional extras: uv sync --extra storage (Zarr on-disk storage), --extra docs (build docs locally), --all-extras. Supports Python 3.10-3.12.
import selectsim as ss
# M, sample_class, alteration_class: see the Introduction tutorial for how
# to build these from a MAF, or load the bundled TCGA LUAD example data.
result = ss.selectX(
M, sample_class, alteration_class,
n_cores=1, min_freq=10, n_permut=1000,
)
# Significant evolutionary dependencies
significant = result['result'][result['result']['FDR']]See the Introduction tutorial for a full walkthrough on real TCGA LUAD data, and docs/ for the full API reference (selectsim.gam for MAF ingestion, selectsim.plotting for visualization, selectsim.io for Parquet/Zarr storage).
If you use SelectSim in your research, please cite:
Iyer A, Petrovic M, Sesia D, Nanni L, Mina M, Ciriello G (2026). Evolving patterns of co-mutations from tumor initiation to metastatic progression. Nature Genetics. DOI: 10.1038/s41588-026-02661-4
Citation metadata is also available in CITATION.cff.
MIT License. See LICENSE.
For bugs or feature requests, use the issue tracker. This is the Python implementation of the SelectSim R package.