Machine Learning Algorithm Identifies an Antibiotic Vocabulary for Permeating Gram-Negative Bacteria
Problem
Many potential antibiotics cannot enter Gram-negative bacteria such as multidrug-resistant Pseudomonas aeruginosa, limiting treatment options.
Why it matters
Multidrug-resistant P. aeruginosa infections are hard to treat, making compounds that can penetrate the bacterium urgent.
What the AI did
A machine learning algorithm identified submolecular fragments ('chemical vocabulary') that promote compound entry into P. aeruginosa, guiding selection of permeating candidates from an external library.
Results
not stated in source
Impact
Identified permeation-promoting fragments and candidate compounds as proof of concept, aiding rational antibiotic discovery.
Record
| Venue | Journal of Chemical Information and Modeling |
|---|---|
| Published | 2020 |
| Work type | article |
| Authors | Rachael A. Mansbach, Inga V. Leus, Jitender Mehla, César A. López, John K. Walker, Valentin V. Rybenkov, Nicolas Hengartner, Helen I. Zgurskaya et al. |
| Citations | 34 (OpenAlex, at import time) |
| References | 60 |
| Model or method | not stated in source |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1021/acs.jcim.0c00352 |
| Open access | green |
| Keywords | machine learning, antibiotic discovery, Gram-negative bacteria, permeability, Pseudomonas aeruginosa |
Importance rating
major — millions of lives, an industry or a field
Abstract calls drug discovery a crisis and multidrug-resistant P. aeruginosa a pressing concern; AMR limits treatment options for a major pathogen, but no scale is quantified.