Figure 2.

Intragenic histone modification regions reflect gene architecture. (a) H3K79me2, and H3K36me3 ChIP-seq reads plotted on two genes (UGCGL1 and SMURF1) of similar length, but different gene architectures. A black line in each density plot shows where a uniform distribution of sequence reads would be. The proposed 5′ intronic, and 3′ exonic histone modification zones are diagrammed below each gene. (b) Heatmaps show the density for H3K4me3, nucleosomes (Nucs), H3K79me2, and H3K36me3 ChIP- and MNase-seq, plotted with respect to transcription start sites (bent arrows) with genes sorted by distances to the beginning of the first internal exons (i.e. first 3′ splice sites). For clarity only the 4286 annotated RefSeq genes that are in the top 50% of expression and with the 50% longest distances to the first internal exon are shown. Full plots are shown in Supplementary Figure 2. Each gene is displayed as a row and columns are 1-kb bins from 10 kb upstream to 50 kb downstream of transcription start sites (bottom). The value of each bin is shaded by the number of ChIP- or MNase-seq reads per kb, scaled to cover data between 0 reads (white) and the 99th percentile of each sequencing experiment (black), indicated at the top right of each heatmap. At far right the aligned transcription start sites (broken line) and the positions of first internal exons (gray line) are shown as a visual reference.