Bayesian Probit Multi-Study Non-negative Matrix Factorization for Mutational Signatures
Blake Hansen, Isabella N. Grabski, Giovanni Parmigiani, Roberta De Vito · 2025
Read Bayesian Probit Multi-Study Non-negative Matrix Factorization for Mutational Signatures by Blake Hansen, Isabella N. Grabski, Giovanni Parmigiani, Roberta De Vito (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-01468.txt.
Description
Blake Hansen and co-authors built a Bayesian method in 2025 for reading the mutational signatures written into cancer genomes. Each mutagenic process, sunlight, tobacco smoke, faulty repair, leaves a characteristic pattern of mutations, and decomposing a tumor's catalog into those patterns reveals its history. Tumors from varied cancer types are more and more collected, but most numeric tools still study one study at a time.
Their multi-study non-negative grid factorization pools several cancer types in one principled model. A mixture form inside it enforces thinness in a smart way: each tumor is explained by a subset of signatures rather than a smear of all of them. And which subset is plausible depends on the patient's covariates. The Bayesian probit form quantifies doubt throughout. The authors argue the design sharpens identification of the signatures themselves while respecting how tumors differ.
Catalog record
| CREATOR | Blake Hansen, Isabella N. Grabski, Giovanni Parmigiani, Roberta De Vito |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 77 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-01468