MCMFPP: A Multifunctional Peptides Prediction Method Based on Class Feature Enhancement and Classifier Fusion

Area of Science:

• * Computational biology and bioinformatics.
• * Drug discovery and development.
• * Peptide science and therapeutics.

Background:

• * Traditional wet-lab methods for peptide function prediction are time-consuming and costly.
• * Existing computational tools struggle with data sparsity, imbalanced data, and feature representation for multifunctional therapeutic peptides (MFTP).
• * Accurate MFTP prediction is essential for developing targeted peptide drugs.

Purpose of the Study:

• * To develop an advanced computational method for accurate multifunctional peptide prediction.
• * To address limitations of existing methods, including data sparsity and inadequate feature representation.
• * To improve the efficiency and accuracy of identifying candidate peptides for therapeutic applications.

Main Methods:

• * Introduction of two subclassifiers: Sequence Learning Feature Enhancement (SLFE) and Class Feature Enhancement Classifier (CFEC).
• * SLFE utilizes the large language model ESMC to enhance sequence representation, particularly for tail-class data.
• * CFEC improves class feature learning using single-function peptide samples and contrastive learning.
• * Proposal of MCMFPP, a deep learning model integrating SLFE and CFEC predictions via weighted fusion.

Main Results:

• * MCMFPP demonstrates superior performance compared to state-of-the-art methods in multifunctional peptide prediction.
• * Achieved improvements include 3.1% in precision, 2.7% in coverage, 2.8% in accuracy, and 2.7% in absolute true rate.
• * Reduced the absolute false rate by 0.3%.
• * Enhanced prediction accuracy for challenging multifunctional peptide samples.

Conclusions:

• * MCMFPP effectively overcomes the limitations of single-classifier approaches in MFTP prediction.
• * The proposed method offers a valuable tool for efficient and accurate identification of multifunctional therapeutic peptides.
• * This advancement can accelerate the development of novel peptide-targeted drugs.