Fig. 2

H2Aub1 is not a repressive mark per se. a Mean wild-type (wt) expression in the absence (−) or presence (+) of H2Aub1 or H3K27me3 after correction for the presence of the other mark (ANCOVA estimated marginal means, in fpkm). b Percentage of upregulated and downregulated genes in PRC1 mutants among the four categories defined by the presence of H2Aub1 and H3K27me3 and dependency of H3K27me3 on PRC1. Asterisks indicate the |log2FC| cut-off used to call significant deregulation: *≥ 4 or **≥ 1; the p value cut-off was 0.05 for both. Small circles indicate the mean of the four mutants. c wt expression level for genes defined by the presence of H3K27me3 and PRC1 dependency. d Expression change in a PRC2 (clf swn) mutant compared to wt for genes defined by the presence of H3K27me3 and PRC1 dependency. e Metagene profiles of H3K27me3 and H2Aub1 in PRC1-dependent and PRC1-independent gene categories. Based on re-analysed ChIP-seq data from [6] (a–e) and RNA-seq data from [20] (a, b*) and [19] (b**, c, d). An ANCOVA post hoc test was used to test for significance in a, and a Mann-Whitney U test for c and d; ns p ≥ 0.05, *p < 0.05, **p < 0.01,***p < 0.001; Bonferroni correction was applied with m = 6. The biological material underlying this figure is whole plants of 7 DAG (ChIP-seq data), 10 DAG (RNA-seq data in a and b*), and 14 DAG (RNA-seq data in b**, c, d)