Skip to main content
Fig. 2 | Genome Biology

Fig. 2

From: SPARK-X: non-parametric modeling enables scalable and robust detection of spatial expression patterns for large spatial transcriptomic studies

Fig. 2

Analyzing the mouse cerebellum Slide-seq data. A Quantile–quantile plot of the observed −log10(P) from different methods against the expected −log10(P) under the null condition in the permuted Slide-seq data. P values were combined across ten permutation replicates. Compared methods include SPARK-X (red), SPARK-G (green) and SpatialDE (steel blue). B Power plot shows the number of genes with spatial expression pattern (y-axis) identified by different methods at a range of FDRs (x-axis) in the Slide-seq data. C Bar plots show the computation time and RAM usage of different methods for analyzing the mouse cerebellum Slide-seq data. D An illustration of the mouse cerebellum, where a cross-section shows its lobular organization. E Bar plot displays the percentage of SE genes identified by either SPARK-X (red) or SPARK-G (green) or both (orange) that were also validated in two gene lists: one from the Harmonizome database (left) and the other from literature (right; Wizeman et al). F Visualization of three representative SE genes identified only by SPARK-X in the Slide-seq data. The top panel shows in situ hybridization results for the three genes obtained from the Allen Brain Atlas. The bottom panel shows relative gene-expression levels (green, high; antique-white, low), with P values from SPARK-X displayed inside parentheses

Back to article page