Extended Data Fig. 5: eQTL conservation and validation.
From: Genetic regulation of gene expression across multiple tissues in chickens
(a) Correlations of lead variant effects (upper triangle) and cis-h2 (lower triangle) between chickens and mammals. The species tree was retrieved from the Timetree5 (https://timetree.org/). P values of Pearson’s correlations were computed using the asymptotic t approximation. (b) Internal validation of eQTLs. Bars in light blue represent the Spearman correlation coefficients of eQTL effect sizes between validation and discovery groups (left y-axis), while red dots represent the π1 statistic, estimating the replication rate of eQTL between groups (right y-axis). Each of the 15 tissues with over 100 individuals was randomly and evenly split into discovery and validation groups. The tissue color legend is shown on the right. (c) Correlation between eQTL effect sizes (x-axis, n = 2,396) and effect sizes from allele-specific expression (ASE) analysis (y-axis) in the liver. P values were computed via the asymptotic t approximation. (d) The proportion of regulatory variants predicted by DeepSEA (prediction score > 0.7) based on 310 functional profiles in chickens (Table S8). molQTL_set: conditionally independent molQTL across tissues; Random_set: randomly selected variants matched for minor allele frequency (MAF) with molQTL; Background: all 1.5 million tested variants. *** P < 0.001, based on a two-sided Student’s t-test.
