Fig. 2

HPV integration alters the local transcriptome and epigenome. a A 10-kbp genomic window centered on TCGA-BA-A4IH’s HPV integration site. reads per million mapped reads (RPM) RNA expression (left); fragments per million (FPM) chromatin accessibility (right). Red: signal from TCGA-BA-A4IH; blue: signal from each of the 8 other HNSC samples. Vertical dashed red line: integration site. b Same data as (a), but in an expanded 1 Mbp genomic window. The green background shows how the coordinates of (a) fit in (b). The purple vertical bars show position of all ATAC-seq peaks found in any of the 9 tumor samples. c (Top): Mapping of genomic positions for peaks with outlier signal in TCGA-BA-A4IH (gray), the position of the HPV integration site (red), and each 250,000 bp tick mark to ATAC-seq peaks. Gray diagonal lines map each 250,000 bp to the corresponding peaks. The black lines map the genomic position of the top 9 peaks with the strongest FPM in any of the 9 samples to the corresponding peaks. (Middle): Heatmap of ATAC-seq peaks in the same 1 Mbp genomic window. Color indicates ATAC-seq FPM divided by the maximum FPM value of chr9 in each patient (see the “ATAC-seq” section). Each column shows a 200-bp genomic window overlapping a peak in any of the 9 patients. We showed all 200-bp genomic windows with sliding windows of 50 bp if the window overlaps a peak. (Bottom): Difference of the values in TCGA-BA-A4IH and the most extreme value in the other 8 patients when TCGA-BA-A4IH had the most extreme value among the 9 patients. We used white when TCGA-BA-A4IH did not have the most extreme value