Fig. 7: Scalability of SV gene identification algorithms.
From: Identification of spatially variable genes with graph cuts
a Memory requirements of scGCO, SPARK, spatialDE, trendSceek, and SOMDE in the number of cells (using 100 genes) on simulated data. b Running time of scGCO, SPARK, spatialDE, trendSceek, and SOMDE in the number of cells (using 100 genes) on simulated data. Dotted line indicates memory or running time extrapolated from measured data. c Running time on adult human heart tissue data with Spatial Transcriptomic sequencing. d Running time on mouse complex tissues data with Slide-seq technology. e Running time on mouse hypothalamus data with over one million cells and hundreds of genes by MERFISH technology. Dotted bars indicate that no results were obtained due to computing errors. Source data are provided as a Source Data file.
