Fig. 4

Eighteen percent of orthologous regulatory elements show significant trans effects. a Schematic depicting the definition of a trans effect: MPRA activity differences between hESCs and mESCs while keeping the sequence constant. b Volcano plot showing the trans effect sizes (log2 fold changes in cell type for human sequences (left) and mouse sequences (right)) of regulatory sequences (black) and null differential controls (gray). Horizontal line depicts an empirical FDR cutoff of 0.1, calculated using null differential controls (see the “Methods” section). c Count of orthologous sequence pairs with significant trans effects (either human or mouse). d Scatter plot showing the trans effects measured for human sequences (x axis) and mouse sequences (y axis) for a given sequence pair, colored by whether they are significant in human (orange), mouse (green), both (black), or neither (gray). Spearman’s rho and number of sequences are shown. e Absolute trans effect sizes across biotypes, broken up into non-conserved TSSs (blue) and conserved TSSs (gray). p values shown are from a two-sided Mann-Whitney test. f Same as d, but this time comparing differences across biotypes. p values shown are from a two-sided Mann-Whitney test