Uncovering evolutionarily remote and highly potent antimicrobial peptides with protein language models - PubMed
4 hours ago
- #antibiotic resistance
- #antimicrobial peptides
- #protein language models
- HMD-AMP is a protein language model-based approach for discovering antimicrobial peptides (AMPs).
- It outperforms previous methods in identifying evolutionarily distant AMPs.
- HMD-AMP was used to predict over 37 million AMPs from host and gut microorganism genomes of nine mammals.
- Experimental validation confirmed 74 out of 91 high-confidence sequences had strong antibacterial activity.
- 48 of these AMPs were evolutionarily remote from known AMPs.
- Four AMPs showed broad-spectrum antibacterial activity at low concentrations and low toxicity.
- The most potent peptide demonstrated therapeutic efficacy in a mouse model of E. coli infection.