SpaGCN: Integrating gene expression, spatial location and histology to identify spatial domains and spatially variable genes by graph convolutional network

By Jian Hu, Xiangjie Li, Kyle Coleman, Amelia Schroeder, Nan Ma, D. Irwin, Eddie B. Lee, Russell T. Shinohara, Mingyao Li. Nature Methods, 2021

962 citations recorded by the source corpus.

Recent advances in spatially resolved transcriptomics (SRT) technologies have enabled comprehensive characterization of gene expression patterns in the context of tissue microenvironment. To elucidate spatial gene expression variation, we present SpaGCN, a graph convolutional network approach that integrates gene expression, spatial location and histology in SRT data analysis. Through graph convolution, SpaGCN aggregates gene expression of each spot from its neighboring spots, which enables the identification of spatial domains with coherent expression and histology. The subsequent domain guided differential expression (DE) analysis then detects genes with enriched expression patterns in the identified domains. Analyzing seven SRT datasets using SpaGCN, we show it can detect genes with much more enriched spatial expression patterns than competing methods. Furthermore, genes detected by SpaGCN are transferrable and can be utilized to study spatial variation of gene expression in other datasets. SpaGCN is computationally fast, platform independent, making it a desirable tool for diverse SRT studies. SpaGCN is a spatially resolved transcriptomics data analysis tool for identifying spatial domains and spatially variable genes using graph convolutional networks.

Subfields: Molecular Biology

View Paper: https://www.semanticscholar.org/paper/81b66947d0fdfb0e05cd21446a1f1bd7010f5f59

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

Citations in the corpus

Spatially Informed Cell Type Deconvolution for Spatial Transcriptomics (2022, background): “Exploring the use of other kernels, such as the periodic kernels 74 , or incorporating histological image information, such as image intensity level, as additional coordinates 6,75 , which can be readily done in CARD, may capture diverse and rich spatial correlation patterns in the future.”
DeepST: identifying spatial domains in spatial transcriptomics by deep learning (2022, background): “SpaGCN (7) combines gene expression, spatial location, and morphology data to identify spatial domains by generating an undirected weighted graph that captures the spatial dependency of the data.”
Spatiotemporal transcriptomic atlas of mouse organogenesis using DNA nanoball patterned arrays (2021, positive): “The distribution of speciﬁc markers ( Pcp4 and Slc17a7 ) captured by Stereo-seq exhibited the expected similarity to in situ hybridization (ISH) data taken from Allen Brain Atlas (ABA) (Lein et al., 2007) and clearer patterns than HDST (Vickovic et al., 2019) or Slide-seqV2-resolved (Stickels et al.”
Unsupervised spatially embedded deep representation of spatial transcriptomics (2021, methodology): “Next, current methods that integrate gene expression with associated spatial information were used to benchmark SEDR, namely SpatialLDA [33], Giotto [26], stLearn [28], SpaGene[34], SpaGCN [29], BayesSpace [27], UTAG [32], DeepST [30], and STAGATE [31] (Additional file 1: Supplementary methods).”
Spatial components of molecular tissue biology (2022, neutral): “They can be broadly categorized into Gaussian process-based frameworks 164,165,168–170 , methods that use hidden Markov random fields 171–173 and graph-based methods 114,174 .”
Graph Representation Learning in Biomedicine and Healthcare (2021, positive): “…graphs (weighted and undirected) and corresponding histopathology images, gene-expression information can be aggregated to generate embeddings of genes that could then be used to investigate spatial domains (to differentiate between cancer and noncancer regions in tissues, for instance) 186 .”
Best practices for single-cell analysis across modalities (2023, methodology): “Our work represents an entry point”