Fig. 3: Common wheat TFBSs underwent one or more round(s) of expansion.

a Fraction of TFBSs in subgenome-homologous regions as determined by the reciprocal alignment across subgenomes. Homoeologou regions syntenic across three subgenomes were defined as homoeo3. b Most enriched GO terms for the subgenome-homoeologous and -specific TFBS-biased targeted genes ranked according to the enrichment P value determined by two-tailed Fisher’s exact test. The upper and lower bars represent high enrichment for subgenome-specific and -homoeologous TFBS targets, respectively. c Box plot showing the distribution of the conservation scores for subgenome-homoeologous (n = 8140) and -specific (n = 14,540) TFBS target genes and all genes (genes in the chrUn were removed, n = 105,200) across four wheat species with different ploidy levels. Horizontal lines in boxplots show median, hinges show interquartile range (IQR), whiskers show 1.5 × IQR, points beyond 1.5 × IQR past hinge are shown. d Circos plots showing subgenome-specific bHLH-1A-1 TFBS pairs with high sequence similarity (BLASTN E-value <1e-30, identity >70%, and query coverage >70%) within each subgenome; the homolog bHLH28 in Arabidopsis thaliana is shown as a control. For each TF peak set, 400 peaks were randomly selected. e Sequence distance distribution between subgenome-specific TFBSs within each subgenome of CS and those of the corresponding TFs in Arabidopsis thaliana (gray line) and Oryza sativa (light blue line). The distance was calculated on the basis of the number of substitutions per 100 bases using the Kimura two-parameter (K2P) model⁷⁹. According to this model, transitions (purine–purine and pyrimidine–pyrimidine) are more likely than transversions (purine–pyrimidine). Gaps are ignored. Source data are provided as a Source Data file.