nnSVG: Scalable Identification Of Spatially Variable Genes Using Nearest-Neighbor Gaussian Processes

nnSVG: Scalable Identification Of Spatially Variable Genes Using Nearest-Neighbor Gaussian Processes


Author(s): Lukas M Weber, Stephanie Hicks

Affiliation(s): Johns Hopkins Bloomberg School of Public Health

Twitter: @lmwebr

Feature selection to identify spatially variable genes (SVGs) is a key step during analyses of spatially resolved transcriptomics data. We introduce 'nnSVG', a scalable new method to identify SVGs based on nearest-neighbor Gaussian processes. Our method can identify SVGs with flexible spatial ranges in expression patterns per gene, can identify SVGs within spatial domains, and scales linearly with the number of spatial locations. We demonstrate the performance of our method using example datasets from the 10x Genomics Visium and Slide-seqV2 platforms and simulations. A software implementation is available from GitHub at https://github.com/lmweber/nnSVG and will be submitted to Bioconductor.