Figure 2

Quality control plots. (a) M-bias plot for the Hansen data, a WGBS experiment on cancer samples. Each sample was sequenced on two flowcells. We show the methylation proportion across each possible read position. This plot shows limited evidence of methylation bias across the read positions. Vertical lines indicate cutoffs used for M-bias filtering. (b) M-bias plots for the Lister data, a WGBS experiment in a fibroblast cell line. These data were aligned using iterative trimming and each read length is depicted separately (different colors). The plot shows methylation bias toward the end of reads for all read lengths. (c) M-bias plot for the Hansen-capture data, a capture bisulfite sequencing experiment on cancer samples. The plot shows methylation bias at the start of the reads.