Abstract:Motivation: Graph embedding learning that aims to automatically learn low-dimensional node representations, has drawn increasing attention in recent years. To date, most recent graph embedding methods are evaluated on social and information networks and are not comprehensively studied on biomedical networks under systematic experiments and analyses. On the other hand, for a variety of biomedical network analysis tasks, traditional techniques such as matrix factorization (which can be seen as a type of graph embedding methods) have shown promising results, and hence there is a need to systematically evaluate the more recent graph embedding methods (e.g. random walk-based and neural network-based) in terms of their usability and potential to further the state-of-the-art. Results: We select 11 representative graph embedding methods and conduct a systematic comparison on 3 important biomedical link prediction tasks: drug-disease association (DDA) prediction, drug–drug interaction (DDI) prediction, protein–protein interaction (PPI) prediction; and 2 node classification tasks: medical term semantic type classification, protein function prediction. Our experimental results demonstrate that the recent graph embedding methods achieve promising results and deserve more attention in the future biomedical graph analysis. Compared with three state-of-the-art methods for DDAs, DDIs and protein function predictions, the recent graph embedding methods achieve competitive performance without using any biological features and the learned embeddings can be treated as complementary representations for the biological features. By summarizing the experimental results, we provide general guidelines for properly selecting graph embedding methods and setting their hyper-parameters for different biomedical tasks. Availability and implementation: As part of our contributions in the paper, we develop an easy-to-use Python package with detailed instructions, BioNEV, available at: https://github.com/xiangyue9607/BioNEV, including all source code and datasets, to facilitate studying various graph embedding methods on biomedical tasks. Contact: yue.149@osu.edu or sun.397@osu.edu Supplementary information: Supplementary data are available at Bioinformatics online.

Survey of Network Embedding for Drug Analysis and Prediction

A Survey of Network Embedding for Drug Analysis and Prediction

A Drug Information Embedding Method Based on Graph Convolution Neural Network.

Network Embedding in Biomedical Data Science

Graph Embedding on Biomedical Networks: Methods, Applications, and Evaluations

Application of network link prediction in drug discovery

Network-Based Methods and Their Applications in Drug Discovery

Network based approach to drug discovery: a mini review.

A Network Embedding Framework Based on Integrating Multiplex Network for Drug Combination Prediction

A survey of drug-target interaction and affinity prediction methods via graph neural networks

Op-cbio190730 1..11

Applying network link prediction in drug discovery: an overview of the literature

Graph convolutional networks for computational drug development and discovery.

Predicting drug-disease associations by network embedding and biomedical data integration.

Predicting combinations of drugs by exploiting graph embedding of heterogeneous networks

A Survey on Graph Embedding Techniques for Biomedical Data: Methods and Applications

Gndd: A Graph Neural Network-Based Method For Drug-Disease Association Prediction

A Node Embedding Framework for Integration of Similarity-based Drug Combination Prediction

Drug Research Meets Network Science: Where Are We?

Comprehensive evaluation of deep and graph learning on drug-drug interactions prediction

Prediction of Drug-target Interactions from Heterogeneous Information Network Based on LINE Embedding Model