PhytoAFP: In Silico Approaches for Designing Plant-Derived Antifungal Peptides

Area of Science:

• Mycology
• Biotechnology
• Computational Biology

Background:

• Emerging infectious diseases (EID) caused by fungi pose significant threats to human health and global food security.
• Plant-derived antifungal peptides (PhytoAFP) represent a promising avenue for therapeutic development against fungal pathogens.

Purpose of the Study:

• To develop and validate a computational model for designing and predicting PhytoAFP.
• To identify key amino acid compositions and motifs associated with PhytoAFP activity.

Main Methods:

• Utilized a support vector machine (SVM) algorithm for model development.
• Analyzed amino acid composition, positional preferences, and motif identification.
• Employed various input features including mono-, di-, and tripeptide composition, and physiochemical properties.

Main Results:

• The model achieved high accuracy, with the TPC-based monopeptide composition model reaching 94.4% accuracy and an MCC of 0.89.
• A dipeptide-based model also demonstrated strong performance with 94.28% accuracy and an MCC of 0.89.
• Identified specific amino acid preferences (C, G, K, R, S) and N-terminal/C-terminal residue patterns.

Conclusions:

• The developed SVM model is effective in predicting PhytoAFP.
• Computational approaches can accelerate the discovery of novel antifungal peptides for therapeutic applications.
• Understanding amino acid composition and motifs is key to designing potent PhytoAFP.