Figure 3From: A statistical method for excluding non-variable CpG sites in high-throughput DNA methylation profilinga - Posterior probability distribution from mixture model. The posterior probability distribution, indicating the likelihood of a probe belonging to the subset of highly correlated, informative probes. b - The number of probes selected at different posterior probability thresholds. The number of probes (y-axis) that will remain at different posterior probability thresholds (x-axis) calculated from the two-class mixture model.Back to article page