【Proposal】A Reinforcement Learning Framework for Protein Function Prediction

Chenxi Hu, Fei Long, Renrui Tian

Keywords: Bioinformatics, deep learning, protein function prediction

Abstract: Proteins are the foundation of all life activities. Accurate predicting protein functionis essential not only for understanding the life at the molecular level but also foradvancing applications in drug discovery and disease treatment. Current methodsuse CNNs and GNNs to extract information from protein sequences and structuresto predict protein function, but we find that these methods often suffer from theinconsistency issue. We will develop a novel reinforcement learning framework toaddress this problem and achieve accurate predictions.

Submission Number: 21