Fig. 4: Genetic analyses to identify regions of the genome responsible for major changes in jaw shape.

All plots are based on 409 LFxTRC F₅ hybrids. a QTL analysis to identify positions in the genome most associated with each trait. b Pleiotropy analysis on linkage group seven to determine whether the oral jaw PC1 trait colocalizes to the same region as the pharyngeal jaw PC1 trait. Significance was determined using a likelihood ratio test (LLRT). c Pleiotropy analysis on linkage group seven to determine whether the oral jaw PC2 trait colocalizes to the same region as the pharyngeal jaw PC1 trait. Significance was determined using a LLRT. d Fine mapping all traits across the entirety of LG7. Values furthest from 0 reflect the largest differences between hybrids with LF and TRC genotypes at a given marker. We find peak genotype-phenotype association at ~50 mb that coincides with our Bayes credible interval (grey bar). Intervals that surround the average phenotypic effect line denote standard error of the mean. e Fine mapping all traits across the Bayes credible interval. Population level genetic diversity (F_ST) data are applied to the map (black dots) with the opacity of each SNP dependent on the degree of segregation between LF and TRC, with those falling above an empirical Z-score threshold of 0.6 determined to be significant, and those above 0.9 deemed highly significant (green lines). Within the credible interval there are four SNPs with F_ST values of 1.0, but a single SNP that falls within a genotype-phenotype peak residing within an intron of dym (black circle). Source data are provided as a Source Data file.