Figure 2.
Method performance. (A) Number of sequencing reads per promoter for all 94 targeted promoters, order by length in base pairs (bp) on the x-axis. Longer promoter amplicons yield fewer sequencing reads (length bias), but 87 amplicons (93%) have coverage within 10-fold of the median coverage (444 reads). The abundance of each promoter ranged from 10 to 5114 reads. (B) Histogram of the pairwise squared correlation coefficients for the number of reads per promoter for all 48 samples. The mean correlation coefficient is 0.91, indicating that the number of reads per promoter is highly reproducible across patient samples.
