Based on computer simulation and experimental verification: mining and characterizing novel antimicrobial peptides...

Area of Science:

• Microbiology
• Bioinformatics
• Food Science

Background:

• Traditional screening for antimicrobial peptides (AMPs) is complex and expensive.
• AMPs are crucial for improving food safety and extending shelf life.
• Novel methods are needed to efficiently discover AMPs from large datasets.

Purpose of the Study:

• To develop and validate a deep learning pipeline for identifying potential AMPs from soil metagenomic data.
• To discover novel AMP candidates with potential applications in food preservation.
• To demonstrate the efficacy of computational approaches in accelerating AMP discovery.

Main Methods:

• A deep learning model was trained and applied to soil metagenomic data.
• Candidate AMPs were screened using surface charge analysis, Hemopred, and ToxinPred.
• Molecular docking and dynamics simulations were used to assess binding affinity and stability.
• Chemical synthesis and antimicrobial activity assays were performed for validation.

Main Results:

• The deep learning pipeline achieved high accuracy (92.71%) and precision (91.29%) in AMP prediction.
• Nine candidate peptides were identified, with P4 (GTAWRWHYRARS) showing the best binding affinity to MrkH.
• Synthesized P4 exhibited significant antimicrobial activity against Klebsiella pneumoniae, Escherichia coli, and Staphylococcus aureus.

Conclusions:

• The developed deep learning pipeline is effective for discovering novel AMPs from metagenomic data.
• Peptide P4 shows potential as a safe and effective antimicrobial agent for food applications.
• This integrated approach accelerates the discovery of AMPs, offering a promising alternative to traditional methods.