De novo design of ligand binding proteins using large language models alone
Nam Hyeong Kim, A. Katherine Hatstat, Hyunil Jo, Yibing Wu, William F. DeGrado · 2026
Read De novo design of ligand binding proteins using large language models alone by Nam Hyeong Kim, A. Katherine Hatstat, Hyunil Jo, Yibing Wu, William F. DeGrado (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/pmc-13684519.txt.
Description
Can a general chatbot design a brand-new protein that grips a chosen target? Nam Hyeong Kim and co-authors put that question to the test in a 2026 study. They asked common large language models to design proteins from scratch, with no copying of known sequences. Targets were metals and greasy small molecules, among them a common pollutant. The models also had to explain the reasoning behind each design.
After shape checks, a small set of designs went to the lab. Metal binders worked in one round, with about a quarter of tries hitting. Pollutant binders took two rounds at a like hit rate. The written reasoning gave judges a way to weigh each design. Some designs drifted from the brief, yet the trial shows real promise.
Catalog record
| CREATOR | Nam Hyeong Kim, A. Katherine Hatstat, Hyunil Jo, Yibing Wu, William F. DeGrado |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 2 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |