Abstract:ABSTRACT Background Link prediction is an important and well-studied problem in network biology. Recently, graph representation learning methods, including Graph Convolutional Network (GCN)-based node embedding have drawn increasing attention in link prediction. Motivation An important component of GCN-based network embedding is the convolution matrix, which is used to propagate features across the network. Existing algorithms use the degree-normalized adjacency matrix for this purpose, as this matrix is closely related to the graph Laplacian, capturing the spectral properties of the network. In parallel, it has been shown that GCNs with a single layer can generate more robust embeddings by reducing the number of parameters. Laplacian-based convolution is not well suited to single-layered GCNs, as it limits the propagation of information to immediate neighbors of a node. Results Capitalizing on the rich literature on unsupervised link prediction, we propose using node similarity-based convolution matrices in GCNs to compute node embeddings for link prediction. We consider eight representative node-similarity measures (Common Neighbors, Jaccard Index, Adamic-Adar, Resource Allocation, Hub- Depressed Index, Hub-Promoted Index, Sorenson Index and Salton Index) for this purpose. We systematically compare the performance of the resulting algorithms against GCNs that use the degree-normalized adjacency matrix for convolution, as well as other link prediction algorithms. In our experiments, we use three-link prediction tasks involving biomedical networks: drug–disease association prediction, drug–drug interaction prediction and protein–protein interaction prediction. Our results show that node similarity-based convolution matrices significantly improve the link prediction performance of GCN-based embeddings. Conclusion As sophisticated machine-learning frameworks are increasingly employed in biological applications, historically well-established methods can be useful in making a head-start. Availability and implementation Our method, SiGraC, is implemented as a Python library and is freely available at https://github.com/mustafaCoskunAgu/SiGraC.

NodeCoder: a graph-based machine learning platform to predict active sites of modeled protein structures

Accurately identifying nucleic-acid-binding sites through geometric graph learning on language model predicted structures

VN-EGNN: E(3)-Equivariant Graph Neural Networks with Virtual Nodes Enhance Protein Binding Site Identification

EquiPocket: an E(3)-Equivariant Geometric Graph Neural Network for Ligand Binding Site Prediction

GGN-GO: geometric graph networks for predicting protein function by multi-scale structure features

A Point Cloud Graph Neural Network for Protein–Ligand Binding Site Prediction

AlphaFold2-aware protein–DNA binding site prediction using graph transformer

Hierarchical, rotation‐equivariant neural networks to select structural models of protein complexes

DeepProSite: Structure-aware Protein Binding Site Prediction Using ESMFold and Pretrained Language Model

OnionNet: a Multiple-Layer Intermolecular-Contact-Based Convolutional Neural Network for Protein–Ligand Binding Affinity Prediction

Bionoi: A Voronoi Diagram-Based Representation of Ligand-Binding Sites in Proteins for Machine Learning Applications

GraphPLBR: Protein-Ligand Binding Residue Prediction with Deep Graph Convolution Network

SGPocket: A New Graph Convolutional Neural Network for Ligand-protein Binding Site Prediction

Functional Protein Structure Annotation Using a Deep Convolutional Generative Adversarial Network

Spatom: a graph neural network for structure-based protein-protein interaction site prediction

RGCNPPIS: A Residual Graph Convolutional Network for Protein-Protein Interaction Site Prediction

PLANET: A Multi-objective Graph Neural Network Model for Protein–Ligand Binding Affinity Prediction

ContactNet: Geometric-Based Deep Learning Model for Predicting Protein-Protein Interactions

State-specific protein-ligand complex structure prediction with a multi-scale deep generative model

Node similarity-based graph convolution for link prediction in biological networks