Fig. 2: Eaf3 is specially required for the normal distribution of Ash1-H3K36me2 occupancy.

a A Venn diagram showing significant overlaps of gene sets among up-regulated genes in the ∆eaf3 and ∆ash1 strains, and Ash1-H3K36me2-marked genes. Significance of overlaps was determined using two-sided Fisher’s exact test and P-values were shown the degree of significance. b-c Box plots and violin plots showing average transcription levels of the genes with and not Ash1-H3K36me2-marked (b) and Ash1-H3K36me3-marked (c) respectively in the indicated strains. For box-plot, the horizontal lines from top to bottom represent the maximum, first quartile, median, third quartile, and minimum of the total data respectively. For b and c, statistical significance was assessed by a two-tailed Student’s test (** P < 0.01). n.s., no significance. FPKM (Fragments Per Kilobase of exon model per Million mapped fragments). d Relative abundance of H3K36me2 and H3K36me3 in the indicated strains. The ratios of H3K36me2 and H3K36me3 to H3 in western blots were calculated by ImageJ software respectively. Values are means ± SD from five biological replicates. Statistical significance to the WT strain was assessed by a two-tailed Student’s test (** P < 0.01). The exact P-values and western blots are shown in the Source Data. e Integrative genome viewers (IGV) showing H3K36me2 and H3K36me3 occupancy in the representative genomic regions of the WT, ∆eaf3, and ∆ash1 strains. f Metaplots and heatmaps showing H3K36me2 ChIP-seq signals in the Ash1-H3K36me2-marked regions of the indicated strains. The average H3K36me2 ChIP-seq signals within 3-kb genomic regions flanking peak summits are shown. CPM (Counts per million). g Metaplots and heatmaps showing H3K36me3 ChIP-seq signals in the Ash1-H3K36me3-marked regions of the indicated strains. For f and g, metaplots and heatmaps are shown with one biological replicate of ChIP-seq assays and similar results were gained in another replicate.