Skip to main content
Fig. 2 | BMC Genomics

Fig. 2

From: BoostMe accurately predicts DNA methylation values in whole-genome bisulfite sequencing of multiple human tissues

Fig. 2

CpG methylation pairwise differences increase with distance and have different average behaviors across chromatin states. Differences were calculated using the average methylation (beta) value of each CpG across all 12 muscle NGT samples. Smoothed lines were fit using a generalized additive model. Average behaviors across chromatin states were similar across all tissue and disease state combinations examined in this work; see Additional file 1: Figure S3, S4. (a) Absolute pairwise differences within chromatin states genome-wide. We randomly sampled blocks of chromatin states genome-wide and used ~ 2 million pairwise differences for each chromatin state. Marginal histograms indicate the number of pairwise differences across chromatin states within the range of the graph. (b) Pairwise differences in blocks of consecutive CpGs with higher across-tissue variance. Variances were calculated across all 58 tissue samples; CpGs with beta value variances above the third quartile of variances were considered as having higher variances. Pairs of CpGs lying in the same high-variance region but different chromatin states were excluded from this analysis

Back to article page