Figure S3.

RNAPII Peaks; Sequence and Enrichment Characteristics, Related to Figure 5

(A) Di-nucleotides frequencies around RNAPII peaks can be explained by mononucleotide frequecies profiles. The y axis shows position specific log-odds ratios between the di-nucleotide frequencies and their expected frequencies based on the mononucleotide frequency profiles. Similar results were seen at greater distances from RNAPII peaks. As expected, CpG di-nucleotides are under-represented in transcribed regions.

(B) RNAPII ChIP-seq peaks in wild-type, shRNA5-, and shRNA7-treated cells overlap. The average, local RNAPII density over peaks called by MACS in the different experiments was calculated in the other experiments as indicated (see color coding). Note that peaks called in wild-type cells (top) on average also had a clear, local increase in RNAPII density in the area upon shRNA5- and shRNA7-mediated RECQL5 depletion. Conversely, peaks called in cells depleted for RECQL5 by shRNA5 (middle) and shRNA7 (bottom) also correlated with an average local increase in RNAPII density in wild-type cells, albeit to a much more modest extent. Method: Each peak region was split into 100 equally sized genomic intervals (bins). The 2 kb region either side of each peak were also split into 100 intervals, together making for 300 genomic intervals centered around the midpoint of each peak. The mean read depth in each bin was calculated for all transcripts and a final mean across all bins was plotted. The red dotted lines delineate the boundaries of the peaks, so from this point of view the x axis is perhaps a little misleading - everything from −150 to −50 is upstream of the peaks, −50 to +50 is the peak itself and +50 to +150 is the region downstream of the peak.