Back

ESKtides: a comprehensive database and mining method for ESKAPE-derived peptides

Li, D. R.; Wu, H.; Zou, G.; Li, X.; Zhang, Y.; Zhou, Y.; Chen, H.; Li, J.

2023-05-04 bioinformatics
10.1101/2023.05.03.538019 bioRxiv
Show abstract

Under the concept of one health, Super-drug resistant bacteria are getting more and more attention from scientists, in particular, ESKAPE bacteria directly killed more than 1,270,000 people in 2019. Recent studies have proposed that phage PGHs(peptidoglycan hydrolases) and antimicrobial peptides can be the new antibacterial agents against multi-drug resistant bacteria. However, Methods for mining antimicrobial peptides based on phages or phages PGHs are in need to develop. Here, using ESKAPE strains and ESKAPE phages in total of 6,809 samples from National Center for Biotechnology Information (NCBI), PhagesDB, Microbe Versus Phage (MVP) and Virus-Host Database, we systematically identified PGHs across ESKAPE strains prophages and phages, mined peptides and scored peptides based on PGHs. As a result, a total of 1,000 high antibacterial activity peptides and 1,200 medium antibacterial activity peptides were identified. In order to reduce the impact of different methods on comments, we used a unified process to comment ESKAPE strains and phages. In addition, we designed an online tool to predict the peptides antibacterial activity. Calculations of peptides phylogeny, peptides physicochemical property and peptides secondary structure are also included into our online tools. Finally, we developed ESKtides, a user-friendly and intuitive database (http://www.phageonehealth.cn:9000/ESKtides) for data browsing, searching, and downloading. ESKtides will significantly provide a rich peptide library based on ESKAPE strains and phages.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.