Colm Seale

MUSICiAn identifies DNA repair genes by analyzing mutational spectra from CRISPR-Cas9 perturbations without controls, revealing known genes and novel candidates.

The paper uses CRISPR mutational spectra to identify functional roles of genes in DNA repair pathways by analyzing mutational signatures with non-negative matrix factorization.

X-CRISP predicts CRISPR repair outcomes with domain adaptability and interpretability, outperforming existing models and improving with transfer learning across cell lines.

The paper introduces SNMF, a supervised method to learn mutational signatures and predict DNA repair deficiencies, achieving high accuracy in cell lines and predicting DDR status in tumors.

ELISL improves synthetic lethality prediction in cancer by integrating context-specific and context-free data, showing better performance and robustness across 8 cancer types.