Extended Data Fig. 1: Comparisons of genome-wide binding signal between different TFs and technical replicates.

(a) Example 20 kilobase region of the A. thaliana genome shows similar TF-binding profiles of four example A. thaliana TFs and their corresponding orthologs from five other species. Numbers after TF IDs indicate percent amino acid identity compared to the A. thaliana ortholog. (b) All-vs-all genome-wide correlations of DAP-seq binding profiles between TFs from six species. Gray grid lines delineate individual groups of orthologous TFs and colored bars indicate TF families. (c) Correlations between different subsets of DAP-seq. Boxplots are drawn according to standard convention, where the box center represents the median, upper and lower box bounds represent the upper and lower limits of the interquartile range, and whiskers extend to the furthest data point within 1.5x the interquartile range from the box. Numbers in parentheses indicate the number of pairwise comparison data points shown in each boxplot. FRIP = fraction of reads in peaks, which serves as a measure of signal-to-noise in each DAP-seq library and can be used as an estimate of overall performance of the assay. Performance in the assay is variable between different TFs, and also varies slightly between technical replicates of the same TF. The category labeled ‘replicate with closest FRIP’ partially controls for this confounding effect on correlations: for each comparison between an A. thaliana TF and an orthologous TF, one technical replicate of each TF was selected such that the difference in FRIP scores between the pair being compared was minimized. In the category labeled ‘same motif group’, motif-groups refers to groups of A. thaliana TFs that fall within the same TF family and also show highly similar motifs (see Supplementary Table 1).