Mining the UniProtKB/Swiss-Prot database for antimicrobial peptides.

Area of Science:

• Bioinformatics
• Peptide Science
• Drug Discovery

Background:

• Antibiotic resistance is a growing global health concern, necessitating the search for novel antimicrobial agents.
• Antimicrobial peptides (AMPs), part of the innate immune system, are promising alternatives to conventional antibiotics.
• Large-scale AMP discovery from public databases using deep learning remains underexplored.

Purpose of the Study:

• To develop and apply a bioinformatics workflow for discovering novel AMPs from large protein databases.
• To identify potential AMPs with applications in the poultry industry.
• To experimentally validate the antimicrobial activity of newly discovered AMP candidates.

Main Methods:

• A novel AMP mining workflow was developed utilizing the AMPlify prediction tool.
• The workflow was applied to the UniProtKB/Swiss-Prot database, focusing on eukaryotic sequences.
• Putative AMPs were prioritized based on structural similarity to known chicken AMPs, and selected peptides were synthesized and tested.

Main Results:

• The workflow identified 8008 novel putative AMPs from eukaryotic sequences.
• Forty AMP candidates were prioritized for potential poultry applications.
• Thirteen out of 38 synthesized peptides demonstrated antimicrobial activity against Escherichia coli and/or Staphylococcus aureus.

Conclusions:

• The developed AMP mining workflow is effective for discovering novel antimicrobial peptides from large sequence databases.
• The identified AMPs show potential for combating bacterial infections, particularly in the poultry sector.
• This study highlights the utility of bioinformatics and deep learning in accelerating the discovery of novel antimicrobial agents.