A mechanism-guided framework for prioritizing membrane-interaction anti-Vibrio peptides from peptidomics data

Area of Science:

• Biochemistry
• Computational Biology
• Microbiology

Background:

• Antimicrobial peptide (AMP) discovery faces challenges in prioritizing effective candidates from complex mixtures.
• Current methods often lack interpretability and fail to account for target-specific membrane interactions.

Purpose of the Study:

• To present a mechanism-guided framework for prioritizing membrane-interaction antimicrobial peptide candidates.
• To integrate machine learning, physicochemical properties, and experimental data for enhanced AMP selection.

Main Methods:

• Developed a framework combining machine-learning AMP screening, membrane-interaction plausibility (MAP) assessment, and a data-driven ranking function (AIPx).
• Utilized literature-derived physicochemical characteristics for MAP and calibrated AIPx using anti-Vibrio peptide data.
• Employed structural visualization for interpretability and validated the framework in a peptidomics study targeting Vibrio spp.

Main Results:

• The MAP + AIPx framework effectively prioritized biologically plausible antimicrobial peptide candidates.
• AIPx showed a consistent relationship with experimentally observed antibacterial activity against Vibrio spp.
• High antibacterial activity correlated with enrichment of high-ranking peptides, not just AMP abundance.

Conclusions:

• The framework enables interpretable and experimentally actionable antimicrobial peptide candidate selection.
• Facilitates species-oriented prioritization of AMPs by considering target-specific membrane characteristics.
• The approach is extensible and supports mechanism-informed prioritization in antimicrobial peptide discovery.