Fig. 3.

Skewed distribution of the large-sized Pol II footprints around the TSS is associated with the directionality of transcription. a Schematic diagram of quantifying the skewness of the Pol II signals at the TSS. The position of the short-fragment peak was identified and then the painted area (i.e. the read coverage) in 500 bp upstream from the peak (denoted as S₋) and 500 bp downstream from the peak (S₊) was calculated. Skewness is expressed by the logarithmic ratio of S₊ to S₋ in this study. b (left) Distribution of the height of the short-fragment coverage from the 13,452 singleton genes. Genes that had the short-fragment signal above the background (dashed line) were selected to allow robust identification of the short-fragment peak (1865 genes). (right) The position of the short-fragment peak relative to the annotated transcription start site for each gene. The vertical axis indicates the index of 1865 genes. Genes that had the peak position within 500 bp around the annotated transcription start site were further selected (1592 genes). c Distribution of the skewness from 1592 genes. The skewness for the two size classes is plotted for each gene (120–270 bp on the x-axis and 270–440 bp on the y-axis). Four genes that were recurrently identified as skewed, either positively or negatively, are indicated in red. d DNA footprints of RNA polymerase II identified by CUT&RUN in this study, and the nascent transcript-sequencing data from the public database. The alignment is shown near the TSS of the four genes from panel (c). The first four rows in each column show CUT&RUN signals from three size classes (40–120 bp, 120–270 bp and 270–440 bp). The same colour code is used as per Fig. 2a to indicate the size classes. The signals were normalised to 10,000 spike-in reads. The scale of the vertical axis in each panel is indicated in bracket. The second four rows show nascent RNA transcription from Gene Expression Omnibus (GSE60358), which contains four mNET-seq assays. The signal of the plus-strand-nascent transcripts is indicated in light blue, and the minus-strand-transcripts in magenta, regardless of the gene orientation shown at the bottom of each column. The grey dashed line indicates the position of the short-fragment peak. e Heatmaps showing the footprints of RNA polymerase II and the directionality of transcription. The top row contains 227 genes that had the large CUT&RUN fragments directed towards downstream of the short-fragment peak (the first three column, positively skewed), and the bottom row contains 131 genes that had the large CUT&RUN fragments directed towards upstream of the short-fragment peak (the first three column, negatively skewed). Each gene is oriented from left to right in the heatmaps (i.e. genes encoded on the minus strand are all flipped horizontally). The heatmaps cover ± 750 bp around the short-fragment peak of each gene. The fourth column onwards shows the mNET-seq datasets from GSE60358. Each of the four datasets contains two heatmaps showing the nascent transcription from either the same strand (sense) or the opposite strand (antisense) of the gene