UBS-seq generated more evenly distributed coverage along the genome. The IQR was used to represent the statistical variance of the data. The box shows the IQR of the data, with a line at the median. The whiskers extend from the box to the 1.5x IQR of the data. When compared to conventional BS-seq, the UBS-seq method showed a 13% and 35% decrease in IQR value for 10 ng and 1 ng samples, respectively. Levene test for equality of variances yielded P values of 0.01 and 2.2 x 10 -32 for 10 ng and 1 ng samples, respectively. IQR, interquartile range.
