A Comparative Analysis of GCN and GAT for Predicting Protein-Protein Interactions

Protein-protein interactions (PPIs) play a fundamental role in biological activities. However, accurately predicting PPIs remains challenging due to the complexity of protein interfaces and the vast number of potential interactions. To address these challenges, this study explores the use of graph attention networks (GATs) and graph convolutional networks (GCNs) to predict PPIs. These deep learning architectures excel at handling graph-structured data, representing proteins as nodes and their interactions as edges. By leveraging various data sources, including sequence and structural data, GCNs and GATs aim to improve prediction accuracy. Ultimately, these advancements hold significant implications for drug discovery, vaccine development, evolutionary biology, and protein function annotation. The curated PPI Dataset can be found at: https://github.com/PrayujS/PPI-Prediction-using-Graph-Nets