InfoSEM: A Deep Generative Model with Informative Priors for Gene Regulatory Network Inference
Tianyu Cui, Song-Jun Xu, Artem Moskalev, Shuwei Li, Tommaso Mansi, Mangal Prakash, Rui Liao · 2025
Read InfoSEM: A Deep Generative Model with Informative Priors for Gene Regulatory Network Inference by Tianyu Cui, Song-Jun Xu, Artem Moskalev, Shuwei Li, Tommaso Mansi, Mangal Prakash, Rui Liao (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-2503-04483.txt.
Description
Tianyu Cui, Song-Jun Xu, Artem Moskalev, and a wider team wrote this machine learning paper, published in 2025. It infers gene regulatory networks, the wiring diagram of which genes switch which others on, without the labeled examples that bias today's best models.
Supervised models ace this task on paper but need costly ground truth and learn the labels' quirks instead of biology. InfoSEM is an unsupervised generative model fed a different clue: text embeddings of gene descriptions, used as informative priors about each gene. Labels can join as an extra prior when available, without importing their biases. The authors also propose a benchmark closer to real biomarker hunting, which exposes rivals' learned biases. Across four datasets, InfoSEM beat existing models by nearly forty percent, and labeled priors added another eleven.
Catalog record
| CREATOR | Tianyu Cui, Song-Jun Xu, Artem Moskalev, Shuwei Li, Tommaso Mansi, Mangal Prakash, Rui Liao |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 57 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-04483