Identification of Membrane Protein Types via Deep Residual Hypergraph Neural Network

Membrane protein's functions are significantly associated with its type. So it is crucial to identify the types of membrane proteins. Conventional computational methods for identifying the species of membrane proteins tend to ignore two issues: high-order correlation among membrane proteins and the scenarios of multi-modal representations of membrane proteins, which leads to information loss. To tackle those two issues, we propose a deep residual hypergraph neural network DRHGNN which enhances the hypergraph neural network with initial residual and identity mapping in this paper. We carry out extensive experiments on four benchmark datasets of membrane proteins. In the meantime, we compare DRHGNN with recently developed advanced methods. Experimental results show the better performance of DRHGNN on membrane protein classification task on four datasets. Experiments also show that DRHGNN can handle the over-smoothing issue with the increase of the number of model layers compared with hypergraph neural network HGNN .