1. A deep learning pipeline was developed to identify antimicrobial peptides (AMPs) from the human gut microbiome.
2. Of 2,349 sequences identified as candidate AMPs, 181 showed antimicrobial activity with a positive rate of >83%.
3. The 11 most potent AMPs were characterized and demonstrated significant efficacy in lowering bacterial load by more than tenfold against a mouse model of bacterial lung infection.
The article is generally reliable and trustworthy, providing evidence for its claims and exploring potential risks associated with the use of the identified AMPs. The authors have provided detailed information on their methodology, including the use of multiple natural language processing neural network models such as LSTM, Attention and BERT to form a unified pipeline for candidate AMP identification from human gut microbiome data. Furthermore, they have also provided evidence for the efficacy of the 11 most potent AMPs in lowering bacterial load by more than tenfold against a mouse model of bacterial lung infection.
However, there are some points that could be further explored in order to increase the trustworthiness and reliability of the article. For example, it would be beneficial to provide more information on how these peptides interact with other components of the human gut microbiome or how they may affect other organisms in an ecosystem context. Additionally, it would be useful to explore potential side effects or risks associated with using these peptides as well as any potential implications for long-term use or misuse. Finally, it would be beneficial to provide more information on how these peptides can be used in clinical settings or what further research is needed before they can be used safely and effectively in humans.