GCNG: graph convolutional networks for inferring gene interaction from spatial transcriptomics data

TLDR

Existing gene‑gene interaction inference methods focus on intracellular interactions, but spatial transcriptomics data enable inference of both intra‑ and inter‑cellular interactions. The authors aim to develop Graph Convolutional Neural networks for Genes (GCNG) to infer gene interactions from spatial transcriptomics data. GCNG encodes spatial relationships as a graph and integrates them with expression data via supervised training, producing outputs usable for downstream functional gene assignment. GCNG outperforms prior spatial transcriptomics methods and identifies novel extracellular gene‑gene interaction pairs. Software and data are available at https://github.com/xiaoyeye/GCNG.

Abstract

Abstract Most methods for inferring gene-gene interactions from expression data focus on intracellular interactions. The availability of high-throughput spatial expression data opens the door to methods that can infer such interactions both within and between cells. To achieve this, we developed Graph Convolutional Neural networks for Genes (GCNG). GCNG encodes the spatial information as a graph and combines it with expression data using supervised training. GCNG improves upon prior methods used to analyze spatial transcriptomics data and can propose novel pairs of extracellular interacting genes. The output of GCNG can also be used for downstream analysis including functional gene assignment. Supporting website with software and data: https://github.com/xiaoyeye/GCNG .