Abstract:Viral infectious diseases are threatening human health and global security by rapid transmission and severe fatalities. The receptor-binding is the first step of viral infection. Identifying hidden virus-receptor interactions opens new perspectives to understand the virus-receptor interaction mechanisms, and further help develop an economical and effective treatment of viral infectious diseases. As a cost-effective strategy, computational methods are adopted to predict potential virus-receptor interactions. However, a small amount of missing values in the raw biological data and zero rows and columns in the virus-receptor interaction matrix affect the stability and precision of the potential virus-receptor interaction prediction method. Therefore, filling out these missing values and zero rows and columns is imperative. In this study, a novel predictive model (PreVRIs) is proposed to predict hidden virus-receptor interactions. In PreVRIs, the viral protein sequence similarity matrix and the viral genomic sequence similarity matrix are computed based on viral protein sequences downloaded from the UniProt database and viral RefSeq genomes obtained from the Reference sequence (RefSeq) database, respectively. We apply the gaussian radial basis function (GRB) to fill in the missing values in the viral protein sequence similarity matrix and the viral genomic sequence similarity matrix, respectively. They are fused into an integrated viral similarity kernel by the similarity kernel learning method (SKL). Second, we calculate the receptor sequence similarity and the receptor protein-protein interaction network similarity based on the amino acid sequence information and the human protein-protein interaction network. Missing values of the receptor protein-protein interaction network similarity matrix are also filled in with GRB. They are also combined into an integrated receptor similarity kernel by SKL. Finally, we fill in zero rows/columns in the interaction matrix with the K-nearest neighbor (KNN) preprocessing method, respectively, and use a matrix completion model to predict hidden virus-receptor interactions. On the viralReceptor sup dataset and the viralReceptor dataset, 10-fold Cross-Validation (10-fold CV) experimental results show that area under curve (AUC) values of PreVRIs are 0.9106 and 0.9252, respectively, which is consistently better than the other related models. In addition, a case study also confirms the effectiveness of PreVRIs to predict hidden virus-receptor interactions.

CBIL-VHPLI: a model for predicting viral-host protein-lncRNA interactions based on machine learning and transfer learning

A Hybrid Prediction Method for Plant lncRNA-Protein Interaction.

A Hybrid Deep Learning Framework for Predicting the Protein-Protein Interaction Between Virus and Host

Transfer Learning Via Multi-Scale Convolutional Neural Layers for Human-Virus Protein-Protein Interaction Prediction

IILLS: Predicting Virus-Receptor Interactions Based on Similarity and Semi-Supervised Learning

Transfer Learning Via Multi-Scale Convolutional Neural Layers for Human–virus Protein–protein Interaction Prediction

A multitask transfer learning framework for the prediction of virus-human protein-protein interactions

Deep Learning-Powered Prediction of Human-Virus Protein-Protein Interactions

Prediction of Virus-Receptor Interactions Based on Multi-View Learning and Link Prediction

ECA-PHV: Predicting Human-Virus Protein-Protein Interactions Through an Interpretable Model of Effective Channel Attention Mechanism

HLPI-Ensemble: Prediction of human lncRNA-protein interactions based on ensemble strategy

Predict the Protein-protein Interaction between Virus and Host through Hybrid Deep Neural Network

DeepViral: prediction of novel virus–host interactions from protein sequences and infectious disease phenotypes

Machine Learning Approaches for Predicting Virus-Human Protein-Protein Interactions: An Evaluation of Retroviral Interaction Networks

Prediction of Virus-Receptor Interactions Based on Similarity and Matrix Completion

A path-based computational model for long non-coding RNA-protein interaction prediction.

Multi-scale Convolutional Neural Networks for the Prediction of Human-virus Protein Interactions

Prediction and Analysis of Human-Herpes Simplex Virus Type 1 Protein-Protein Interactions by Integrating Multiple Methods

Learning Distributed Representations of RNA and Protein Sequences and Its Application for Predicting Lncrna-Protein Interactions

Prediction of Human-Virus Protein-Protein Interactions Through a Sequence Embedding-Based Machine Learning Method

Prediction of Plant lncRNA-Protein Interactions Using Sequence Information Based on Deep Learning