Gene Regulatory Network Inference through Link Prediction using Graph Neural Network

Gene Regulatory Networks (GRNs) depict the causal regulatory interactions between transcription factors (TFs) and their target genes [2], where TFs are proteins that regulate gene transcription. GRN plays a vital role in explaining gene function, which helps to identify and prioritize the candidate genes for functional analysis [3]. Currently, high-dimensional transcriptome datasets are produced from high-throughput sequencing techniques, such as microarray and RNA-Seq. These techniques can capture the differences in the expression of thousands of genes at once. Through these wet-lab experiments, studying the interconnections among a large number of genes or TFs at a network level is challenging [4]. Therefore, one of the important topics in computational biology is the inference of GRNs from high-dimensional gene expression data through statistical and machine learning approaches [2].