Antibiotic discovery with artificial intelligence for the treatment of Acinetobacter baumannii infections
Problem
Multidrug-resistant Acinetobacter baumannii infections are difficult to treat, and new antibiotic candidates are needed against targets such as outer membrane protein W (OmpW).
Why it matters
Drug-resistant A. baumannii infections leave clinicians with few treatment options.
What the AI did
QSAR machine-learning models trained on a ChEMBL set of >7,000 compounds with MIC values screened a library of 11,648 natural compounds, followed by structure-based virtual screening against OmpW and in silico pharmacokinetic evaluation.
Results
Top ten compounds bound OmpW with energy scores of −7.8 to −7.0 kcal/mol, mostly curcuminoids; lead compound validation outcome not stated (abstract truncated).
Impact
Identified curcuminoid lead compounds as potential antibiotic candidates; no demonstrated treatment efficacy in the evidence.
Record
| Venue | mSystems |
|---|---|
| Published | 2024 |
| Work type | article |
| Authors | Yassir Boulaamane, Irene Molina Panadero, Abdelkrim Hmadcha, Celia Atalaya Rey, Soukayna Baammi, Achraf El Allali, Amal Maurady, Younes Smani |
| Citations | 33 (OpenAlex, at import time) |
| References | 53 |
| Model or method | QSAR models (no specific algorithm named) |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1128/msystems.00325-24 |
| Open access | gold |
| Keywords | Acinetobacter baumannii, antibiotic resistance, QSAR, virtual screening, drug discovery |
Importance rating
significant — thousands to millions affected
Abstract calls multidrug-resistant A. baumannii a global challenge with few treatment options; new antibiotic candidates address a major AMR threat, though only in silico/in vitro evidence shown.