Contextualizing biological perturbation experiments through language
Menghua Wu, Russell Littman, Jacob Levine, Lin Qiu, David Richmond, Tommaso Biancalani, Jan-Christian Hütter · 2025
Read Contextualizing biological perturbation experiments through language by Menghua Wu, Russell Littman, Jacob Levine, Lin Qiu, David Richmond, Tommaso Biancalani, Jan-Christian Hütter (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-2502-21290.txt.
Description
Menghua Wu and colleagues use language models to reason over cell perturbation screens, in work from 2025. Rich perturbation tests probe biology finely, but cost bars their wide use today. Learning tools ignore the rich biology written in papers and gene lore so far.
Their PerturbQA benchmark asks models to predict unseen perturbation effects from context. Tasks cover expression change, direction of change, and gene set enrichment calls. Leading learning and stats tools score poorly on the new benchmark overall. A simple read, fetch, and answer language frame matches or beats those tools. Language thus offers a workable medium for biological experiment logic.
Catalog record
| CREATOR | Menghua Wu, Russell Littman, Jacob Levine, Lin Qiu, David Richmond, Tommaso Biancalani, Jan-Christian Hütter |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 83 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-21290