Product category

Finding Peptide by Machine Learning?

Yes. The latest Nature Bioengineering has one article describing how to use machine learning to find potent antimicrobial peptides.

Finding functional peptides through systematic identification is challenging due to the large number of possible peptide sequences. A machine-learning pipeline has been developed to search through billions of sequences in a virtual library of peptides made of 6-9 amino acids to locate potent antimicrobial peptides. The pipeline includes trainable machine-learning modules that are arranged in a coarse-to-fine design to gradually reduce the search space. The top three antimicrobial hexapeptides identified by the pipeline were found to be effective against a variety of drug-resistant pathogens. In tests on mice with bacterial pneumonia, the identified peptides were as effective as penicillin, with low toxicity and a low risk of inducing drug resistance. This machine-learning pipeline may speed up the discovery of new functional peptides.