Antimicrobial peptides recognition using weighted physicochemical property encoding.

Area of Science:

• Computational biology
• Biochemistry
• Infectious disease research

Background:

• Antimicrobial resistance (AMR) poses a significant global health threat, necessitating novel therapeutic strategies.
• Antimicrobial peptides (AMPs) are vital host defense components effective against multidrug-resistant pathogens.
• Current AMP screening methods are costly and time-intensive, highlighting the need for efficient computational tools.

Purpose of the Study:

• To develop precise and rapid computer-aided models for the preliminary selection of antimicrobial peptides (AMPs).
• To introduce a novel peptide encoding technique, amino acid index weight (AAIW), for enhancing AMP recognition.
• To create specialized models for identifying antimicrobial, antibacterial, antiviral, and antifungal peptides.

Main Methods:

• Trained four AMPs recognition models (antimicrobial, antibacterial, antiviral, antifungal) utilizing the amino acid index weight (AAIW) encoding method.
• Integrated datasets from DRAMP and other published databases for model training.
• Evaluated model performance on two independent test sets.

Main Results:

• The developed AAIW-based AMPs recognition models demonstrated high performance, surpassing previous methods.
• All four models achieved over 93% accuracy and a Matthew's correlation coefficient (MCC) of 0.87.
• An accessible online server for AMP recognition is available at https://amppred-aaiw.com.

Conclusions:

• The AAIW encoding method provides a robust foundation for developing accurate computational tools for AMP identification.
• The proposed models offer a significant advancement in accelerating the discovery of novel antimicrobial peptides.
• This research facilitates faster preliminary screening of AMPs, reducing laboratory experimental costs and time.