Fig. 5
![Fig. 5](http://media.springernature.com/full/springer-static/image/art%3A10.1186%2Fs12864-022-09082-7/MediaObjects/12864_2022_9082_Fig5_HTML.png)
The relationship between Grh binding, chromatin accessibility, and gene expression variation. A Region/gene pairs were grouped based on evidence of a hybrid allelic difference (q < 0.05) for the indicated combination of data types (below plot). Evidence for and no evidence for a difference between the hybrid alleles is indicated by triangles and equal signs, respectively. For each group, the observed number of region/gene pairs is plotted as a red point, and the gray violin plots represent the distribution of counts obtained from 1000 permutations in which the rows (representing region/gene pairs) were shuffled for each datatype. From these permuted distributions, we calculated an empirical p value: *** < 0.001, * p < 0.05, ***, and n.s. = non-significant. B For three groups of interest, spaghetti plots of the absolute value of the estimated gene expression difference between hybrid alleles for Grh binding (Grh), chromatin accessibility (Acc.), and gene expression (Exp.). Loess lines (blue) were fit for each group to summarize the trends. For the molecular models, greater Grh binding, chromatin accessibility, and gene expression in Zhr is solely for the purposes of example; in some cases, the Z30 allele is expressed more strongly than the Zhr allele