Fig. 2: Interspecific constraint of maize ACRs in grasses.

a The proportion of constrained peptides (blue), genes (purple), ACRs (green) and randomly selected sequences (red) with the increasing of pairwise divergence time (Mya) between maize and other non-maize genomes. The error bands indicate 95% confidence intervals. b Distribution of maize ACRs according to their constraint in other grass genomes. N1 denotes the number of grass species in which ≥ 90% of maize ACRs can be aligned. N2 denotes the number of grass species in which ≤ 10% of maize ACRs can be aligned. Three distinct groups (highly constrained, moderately constrained and maize specific) are highlighted in squares. c The distribution of genomic features of six groups including highly constrained, moderately constrained, maize specific, class I maize specific, class II maize specific ACRs and randomly selected ACRs as control. d Barplot showing the composition of different TE superfamilies in different categories of ACRs as shown in (c). e UMAP projection of all 80,365 ACRs based on their constraint level in each of the non-maize genomes. Each point is one ACR and the colors represent ACR groups. f The chromatin accessibility of ra1 (Zm00001d020430) and its ~40 kb flanking region. The red region indicates a maize specific ACR. The structure of genes is shown in the bottom track. g Tissues specificity of genes targeted by highly constrained ACRs and maize-specific ACRs across 12 tissues. The P-values were calculated by two-sided Wilcoxon test. Tau was used as the index of tissue specificity. The index varied from 0 (consistently expressed genes) to 1 (strongest tissue-specificity of gene expression). The box denotes the maxima, 25th, median, 75th percentiles and minima, and the whiskers indicate the 1.5× interquartile range. Source data are provided as a Source Data file.