HCNS:A deep learning model for identifying essential proteins based on hypergraph convolution and sequence features

Accurate identification of essential proteins is crucial in biomedical research. Traditional methods rely on Protein-Protein Interaction (PPI) networks and limited biological feature data, often neglecting the critical role of protein amino acid sequences. To address this, we propose the HCNS model, which integrates the Hypergraph Convolutional Network (HGCN) module, the Seq-CNN-MB-NAG feature extraction module, and the Multi-Layer Perceptron (MLP) recognition module, significantly enhancing the accuracy of essential protein identification. The HGCN module constructs a hypergraph from the PPI network and protein complex data, capturing the complex structural relationships between proteins. The Seq-CNN-MB-NAG module utilizes Convolutional Neural Networks (CNN), Multi-Head Self-Attention (MHSA), Bidirectional Long Short-Term Memory (Bi-LSTM), and NAG Transformer to process protein amino acid sequences and extract sequence features. The MLP module then fuses these two sets of features for precise recognition. Experimental results show that the HCNS model outperforms existing methods, achieving an accuracy of 93.38%, with an Area Under the Curve (AUC) of 98.33% and an Area Under the Precision-Recall Curve (AUPR) of 97.16%, demonstrating its potential in essential protein identification.