Embedded-AMP: A Multi-Thread Computational Method for the Systematic Identification of Antimicrobial Peptides...

Area of Science:

• Biochemistry
• Computational Biology
• Drug Discovery

Background:

• Antimicrobial peptides (AMPs) are promising alternatives to conventional antibiotics for combating resistance.
• Current AMP identification methods are limited, often overlooking AMPs embedded within larger proteins.

Purpose of the Study:

• To develop a machine learning (ML)-based pipeline for identifying AMPs embedded within proteomes.
• To overcome the limitations of traditional AMP discovery methods.

Main Methods:

• In-silico digestion of proteomes to generate k-mers of varying lengths.
• Computation of molecular descriptors for each k-mer.
• Antimicrobial activity prediction using ML models.

Main Results:

• The pipeline efficiently identified AMPs in the shrimp proteome within 20 minutes.
• Analysis of rodent proteomes revealed a positive correlation between species longevity and the abundance of predicted AMPs.
• The ML pipeline demonstrated high efficiency in predicting embedded AMPs.

Conclusions:

• The proposed ML pipeline effectively identifies embedded antimicrobial peptides (AMPs) within proteomes.
• The findings suggest a potential link between AMPs and species longevity.
• The pipeline offers a valuable tool for discovering novel AMPs for therapeutic applications.