Abstract:We present a novel formulation of graph classification in inferring gene regulatory relationships from gene expression and graph embedding.Our method leverages a powerful framework, gene regulatory graph neural network (GRGNN), which is flexible and powerful to ensemble statistical powers from a number of heuristic skeletons.Our results show GRGRNN outperforms previous supervised and unsupervised methods inductively on benchmarks.GRGNN can be interpreted and explained following the biological network motif hypothesis in gene regulatory networks.Discovering gene regulatory relationships and reconstructing gene regulatory networks (GRN) based on gene expression data is a classical, long-standing computational challenge in bioinformatics. Computationally inferring a possible regulatory relationship between two genes can be formulated as a link prediction problem between two nodes in a graph. Graph neural network (GNN) provides an opportunity to construct GRN by integrating topological neighbor propagation through the whole gene network. We propose an end-to-end gene regulatory graph neural network (GRGNN) approach to reconstruct GRNs from scratch utilizing the gene expression data, in both a supervised and a semi-supervised framework. To get better inductive generalization capability, GRN inference is formulated as a graph classification problem, to distinguish whether a subgraph centered at two nodes contains the link between the two nodes. A linked pair between a transcription factor (TF) and a target gene, and their neighbors are labeled as a positive subgraph, while an unlinked TF and target gene pair and their neighbors are labeled as a negative subgraph. A GNN model is constructed with node features from both explicit gene expression and graph embedding. We demonstrate a noisy starting graph structure built from partial information, such as Pearson's correlation coefficient and mutual information can help guide the GRN inference through an appropriate ensemble technique. Furthermore, a semi-supervised scheme is implemented to increase the quality of the classifier. When compared with established methods, GRGNN achieved state-of-the-art performance on the DREAM5 GRN inference benchmarks. GRGNN is publicly available at https://github.com/juexinwang/GRGNNDownload : Download high-res image (78KB)Download : Download full-size image

Fast and accurate gene regulatory network inference by normalized least squares regression

Optimal Sparsity Selection Based on an Information Criterion for Accurate Gene Regulatory Network Inference

Netreg: Network-Regularized Linear Models for Biological Association Studies

Inference of Gene Regulatory Network Based on Local Bayesian Networks

GMFGRN: a matrix factorization and graph neural network approach for gene regulatory network inference

SIN-KNO: A method of gene regulatory network inference using single-cell transcription and gene knockout data

Accurate inference of gene regulatory interactions from spatial gene expression with deep contrastive learning

Inferring the experimental design for accurate gene regulatory network inference

A Sparse Reconstruction Approach for Identifying Gene Regulatory Networks Using Steady-State Experiment Data

Inferring single-cell gene regulatory network by non-redundant mutual information

A hybrid deep learning framework for gene regulatory network inference from single-cell transcriptomic data

Gene regulatory network inference based on novel ensemble method

Gene regulatory network inference using PLS-based methods

Inference of gene regulatory networks based on nonlinear ordinary differential equations

GRNbenchmark - a web server for benchmarking directed gene regulatory network inference methods

DiscoGen: Learning to Discover Gene Regulatory Networks

Inductive inference of gene regulatory network using supervised and semi-supervised graph neural networks

iLSGRN: inference of large-scale gene regulatory networks based on multi-model fusion

Gene regulation network inference using k-nearest neighbor-based mutual information estimation: revisiting an old DREAM

Gene regulatory network inference using fused LASSO on multiple data sets

Fused Regression for Multi-source Gene Regulatory Network Inference