Target-AMP: Computational prediction of antimicrobial peptides by coupling sequential information with evolutionary...

Area of Science:

• Biochemistry
• Computational Biology
• Machine Learning

Background:

• Antimicrobial peptides (AMPs) show promise in combating antibiotic resistance due to their broad-spectrum activity.
• The rapid discovery of AMPs necessitates efficient computational methods for identification.
• Wet-lab validation of AMPs is costly and time-consuming.

Purpose of the Study:

• To develop an accurate machine learning-based classification system for predicting antimicrobial peptides.
• To create an efficient computational tool to aid in the selection of AMP candidates before in-vitro testing.

Main Methods:

• Feature extraction using Position-Specific Scoring Matrix (PSSM), Pseudo Amino Acid Composition, and di-peptide composition.
• Application of machine learning classifiers: K-nearest neighbor (KNN), Random Forest (RF), and Support Vector Machine (SVM).
• Development of the Target-AMP prediction system.

Main Results:

• The Target-AMP predictor achieved high accuracy rates of 97.07% on the independent dataset and 95.71% on the training dataset.
• The proposed method demonstrated superior performance compared to existing techniques.
• Successful distinction between antimicrobial peptides using integrated feature extraction methods.

Conclusions:

• The developed Target-AMP system provides an effective and accurate computational approach for antimicrobial peptide prediction.
• This tool can significantly accelerate the discovery of novel AMPs for therapeutic applications.
• Target-AMP offers a valuable solution to overcome the limitations of traditional experimental methods in AMP research.