Fig. 2: Power analysis for 2D simulation.
From: Dimension-agnostic and granularity-based spatially variable gene identification using BSP
A Spatial expression patterns I-III (left to right) as defined in SpatialDE and SPARK. B Power comparison among the different methods with moderate noise level (\(\sigma\)=0.5), and moderate signal strengths (3-folds) for the spatial expression patterns I, II, and III (left to right). All simulations were generated based on the mouse olfactory bulb data with 260 spots of cells. Each simulation replicate contains 1000 SVGs and 9000 non-SVGs. C Computational time with an increasing number of genes and fixed 2000 spots. The y-axis indicates the logarithmic of the running time in seconds. D Computational time with an increasing number of spots and fixed 10,000 genes. The y-axis indicates the logarithmic of the running time in seconds. The time greater than 48 hours is not shown in the figure. Source data are provided as a Source Data file.
