VAE-MS: An Asymmetric Variational Autoencoder for Mutational Signature Extraction
Ida Egendal, Rasmus Froberg Brøndum, Dan J Woodcock, Christopher Yau, Martin Bøgsted · 2026
Read VAE-MS: An Asymmetric Variational Autoencoder for Mutational Signature Extraction by Ida Egendal, Rasmus Froberg Brøndum, Dan J Woodcock, Christopher Yau, Martin Bøgsted (2026) 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-2602-22239.txt.
Description
Cancer leaves mutation patterns that reveal the body processes behind a tumor, but reading those patterns well has been hard. Ida Egendal and four co-authors introduce VAE-MS, a new model that pairs a lopsided autoencoder design with probabilistic methods. It aims for truer estimates that better capture how tumors vary.
The model is tested against three top rivals, including the factorization gold standard and two autoencoder variants. By mixing lopsided design with probability, VAE-MS targets the twin flaws of past methods: accuracy and honest uncertainty. Doctors gain a signature tool built for the messy variation of real tumor data.
Catalog record
| CREATOR | Ida Egendal, Rasmus Froberg Brøndum, Dan J Woodcock, Christopher Yau, Martin Bøgsted |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 29 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-22239