A general language model for peptide function identification

TL;DR

PDeepPP is a versatile deep learning framework that leverages pretrained protein language models and hybrid architectures to accurately identify diverse peptide functions and post-translational modifications, advancing biomedical research.

Contribution

This work introduces PDeepPP, a unified deep learning model that improves generalizability and performance in peptide function and PTM identification across multiple biological tasks.

Findings

Achieves state-of-the-art performance in 25 of 33 tasks

High accuracy in antimicrobial and phosphorylation site prediction

Significant reduction in false negatives for antimalarial peptides

Abstract

Accurate identification of bioactive peptides (BPs) and protein post-translational modifications (PTMs) is essential for understanding protein function and advancing therapeutic discovery. However, most computational methods remain limited in their generalizability across diverse peptide functions. Here, we present PDeepPP, a unified deep learning framework that integrates pretrained protein language models with a hybrid transformer-CNN architecture, enabling robust identification across diverse peptide classes and PTM sites. We curated comprehensive benchmark datasets and implemented strategies to address data imbalance, allowing PDeepPP to systematically extract both global and local sequence features. Through extensive analyses including dimensionality reduction and comparison studies, PDeepPP demonstrates strong, interpretable peptide representations and achieves state-of-the-art performance in 25 of the 33 biological identification tasks. Notably, PDeepPP attains high accuracy in antimicrobial (0.9726) and phosphorylation site (0.9984) identification, with 99.5% specificity in glycosylation site prediction and substantial reduction in false negatives in antimalarial tasks. By enabling large-scale, accurate peptide analysis, PDeepPP supports biomedical research and the discovery of novel therapeutic targets for disease treatment. All code, datasets, and pretrained models are publicly available via GitHub (https://github.com/fondress/PDeepPP) and Hugging Face (https://huggingface.co/fondress/PDeppPP)