Supplementary Figure 7: Quality metrics for ChIP-seq analysis of H3K27me3.
(a) Distribution of H3K27me3 peaks is highly reproducible between replicates. (b) Selection of H3K27me3 peaks. Two representative H3K27me3 ChIP-seq experiments (x-axes) are shown relative to input ChIP-seq (y-axis). H3K27me3 ChIP peaks with fewer reads (e.g. those with RPKM < 0.75) are more likely to be caused by high input signal rather than genuine binding of modified histones. A Poisson test was used to test for statistical enrichment of reads in H3K27me3 peaks with P < 10-7 set as a cutoff. (c) ChIP-qPCR validation of common and differential H3K27me3 peaks. Experiment done in triplicate.
