Halicin, an antibiotic found by a neural network screening a compound library
A graph neural network trained on 2,335 molecules picked halicin out of a repurposing library; it killed multidrug-resistant bacteria including Acinetobacter baumannii and Mycobacterium tuberculosis in vitro and cleared infections in mice.
- Model
- Directed message-passing graph neural network (Chemprop)
- Field
- Medicine
- Date
- 2020-02-20
- Human collaborators
- Jonathan Stokes, Regina Barzilay, James Collins
Sources
Original work
Announcement
What was found
The model was trained to predict growth inhibition of E. coli from structure alone, then applied to the Drug Repurposing Hub. Halicin, an abandoned diabetes candidate, scored highly despite being structurally unlike known antibiotics. It kills by dissipating the proton-motive force across the bacterial membrane, and the team could not evolve resistant E. coli over 30 days of serial passage. A follow-up screen of about 107 million molecules from ZINC15 produced eight further candidates. Published in Cell.
Novelty check
Virtual screening long predates deep learning. What was new was a model surfacing a hit structurally distant from its training set, with a mechanism distinct from existing antibiotic classes, and the follow-through to efficacy in animals. Halicin itself was a known molecule; the discovery is of its antibacterial activity, not of the compound.
Caveats and known objections
Halicin had not entered clinical trials six years on, which is the usual fate of antibiotic leads and a reminder that discovery is not the bottleneck in this field. The model ranked compounds; the mechanism work, the animal studies and the interpretation are human. Membrane-disrupting antibacterials carry a known selectivity risk in mammals.
Nobody outside the lab has checked this yet.
Reading the primary source closely enough to say whether it supports the claim counts as a check, and you are credited on the entry.
Or on GitHub: submit a check challenge the grade send a correction or send a pull request
Flag this for triage
Signals order the review queue and nothing else. They are never published, and they never move a grade: that takes a citation.
Entry history (1 event)
- AddedEntered the registry graded Peer reviewed and Search scaffold.
Entries are never deleted. A grade that does not hold up is downgraded on the record, with the reason beside it.
Graded peer reviewed for verification and search scaffold for autonomy. What these mean.
Cite this entry
whataifound.org. (2020). Halicin, an antibiotic found by a neural network screening a compound library. whataifound.org: A Registry of AI Scientific and Mathematical Discoveries. https://whataifound.org/finding/2020-02-20-halicinBibTeX
@misc{whataifound-mit-2020-halicin,
title = {Halicin, an antibiotic found by a neural network screening a compound library},
author = {{whataifound.org}},
year = {2020},
howpublished = {whataifound.org: A Registry of AI Scientific and Mathematical Discoveries},
note = {Result by MIT / Broad Institute. Verification: Peer reviewed. Autonomy: Search scaffold.},
url = {https://whataifound.org/finding/2020-02-20-halicin}
}