PepLM-GNN

Introduction


The precise prediction of peptide-protein interaction (PepPI) is a core support for promoting breakthroughs in peptide drug research, as well as understanding the regulatory mechanisms of biomolecules. Reserachers have developed several computational methods to predict PepPI. However, existing computational methods also have significant limitations.Inthis study, we propose a computing framework, PepLM-GNN, that integrates pre-trained language ProtT5 model with hybrid graph network for accurate identification of PepPI. This model constructs a graph by using ProtT5-extracted semantic context features of peptides and proteins to form heterogeneous nodes, with edges connecting interacting peptide-protein pairs. The hybrid graph network utilizes Graph Convolutional Networks (GCN) to provide the comprehensive information of the peptide and protein sequences, while employing the Graph Isomorphism Network (GIN) to capture the global interations between them.Compared with the existing advanced methods, PepLM-GNN demonsited the high accurately performance and robustness in predicting the PepPIs. We further demonstrated the capabilities of PepLM-GNN in virtual peptide drug screening, which is expected to facilitate the discovery of peptide drugs and the elucidation of protein functions.

PepLM-GNN Model Architecture

Figure 1. The framework of PepLM-GNN, comprising four modules: ProtT5, graph convolution, graph isomorphism, and classification.

Cite

Upon the usage the users are requested to use the following citation:

Ke Yan, Meijing Li, Shutao Chen, Tianyi Liu, and Bin Liu*.
PepLM-GNN: A Graph Neural Network Framework Leveraging Pre-trained Language Models for Peptide-Protein Binding Prediction. (Submitted)