Figure 3—figure supplement 1.

(A) CUT&RUN for X. laevis blastulae requires cell dissociation prior to nuclear extraction. (B) Comparison of different nuclear extraction techniques. Percent DNA recovered was estimated by NanoDrop quantification of phenol-chloroform extracted DNA after nuclear extraction, as a percentage of theoretical total nuclear DNA mass based on the length of the reference genome sequence. Three nuclear extraction methods were tested with and without cell dissociation: gentle washing by pipeting buffer on the surface of the cells, pipet mixing, vortexing at 1500 rpm. (C) Heatmap of pairwise sample correlation between CUT&RUN samples, as measured by log2 coverage in a 1 Kb window around the center of ATAC-seq open regions (N=58223). (D) Heatmaps showing CUT&RUN coverage over transcription start sites for all samples. (E) Boxplots as in Figure 3B showing CUT&RUN log2 coverage L/S on genes in which only one homeolog is activated at stage 9 (L, S) or both activated (LS), stratified into maternal-zygotic genes (MZ) with >1 TPM maternal contribution (N = 547, 889, 398 for L, LS, and S respectively), and strictly zygotic genes (Z) (N = 46, 126, 35 for L, LS, and S respectively). Two-way ANOVA for each of the four sets indicates there is no significant interaction between MZ/Z and homeolog activation pattern (p=0.37–0.91). (F) Biplots comparing log2 L versus S RNA-seq activation ratio (x axis) and log2 L versus S CUT&RUN read coverage. Linear regression lines are overlaid in red; each has positive slope. p Values are from two-sided Pearson’s correlation tests. (G) CUT&RUN coverage over paired homeologous gene regions around the TSS. Gene pairs are sorted according to L versus S RNA-seq activation ratio (right).