Fig. 1

Stochastic cytosine methylation patterns are selectively dependent on CGGBP1 depletion

a A schematic representation of the two main methods used to quantify and compare MeDIP signals from CT and KD. The normalized MeDIP signal was obtained for CT (blue) and KD (red) in 0.2 kb bins of hg38 and the Diff/Sum ([KD-CT]/[KD + CT]) ratios were calculated for each bin pair. The frequency plots of Diff/Sum ratios and MeDIP signals were used to compare CT and KD. b Diff/Sum frequency in HEK293T shows a stochastic distribution resulting in a near-congruent gain-of-methylation (GoM) and loss-of-methylation (LoM). c A comparison of observed distribution and artificially generated expected distribution of methylation signal Diff/Sum ratios (see methods for details; plotted with a frequency interval of 0.1) for HEK293T shows a left-shift towards the negative Diff/Sum values (mean of differences − 0.2486; p value 0.0018). d The widespread GoM and LoM, as shown in b, nullify any net change in cytosine methylation resulting in highly similar methylation frequencies in HEK293T CT and KD. e Representative genome-view of HEK293T CT and KD MeDIP signals. f The Diff/Sum ratio distribution for GM02639 has a skewed distribution showing a net GoM. g Similar to the analysis in c, a comparison of observed and expected distributions for GM02639 showed a right-shift towards the positive Diff/Sum values (mean of differences 0.3609; p value 0.0009). h The MeDIP signal distribution for GM02639 CT and KD show that the GoM and LoM are restricted to regions with low and high methylation levels respectively. i A representative genome browser view of MeDIP signals for CT and KD.