Skip to main content
. 2023 Apr 10;25(2):541–558. doi: 10.1093/biostatistics/kxad007

Fig. 3.

Fig. 3.

Applying SICERS to clinical data Inline graphic (a)–(c) and replication data Inline graphic (d)–(f). (a) A heatmap of Inline graphic of the first data set (Inline graphic); hotter pixels indicate more differential edges between cases and controls, and there is no apparent topological pattern for these hot edges. (b) We then perform SICERS in Inline graphic and find a significant subnetwork [the bold square, which is magnified in (c)]. (c) The enlarged disease-relevant subnetwork in Inline graphic with region names. (d) A heatmap of Inline graphic of the second data set (Inline graphic). (e) The disease-relevant subnetwork was detected by using Inline graphic alone. (f) The enlarged network in Inline graphic with region names. To save space here, versions of the enlarged plots (c) and (f) with more-readable axis labels are included in the supplementary material available at Biostatistics online.