Figure 3

Long-read SMRT-BS and reproducibility. (A) To assess if amplicon length influenced methylation quantitation, CpG islands with previously determined low (TUBGCP3), intermediate (MEST), and high (EHPA8) methylation levels were subjected to SMRT-BS with four overlapping amplicons ranging from 625–1491 bp. Each amplicon was designed to completely or partially cover the CpG island (highlighted by green bars) and each was independently amplified in triplicate to determine reproducibility. (B) A heat map of CpG methylation identified by SMRT-BS for amplicons illustrated in (A), demonstrating strong consistency with methylation quantitation between triplicate amplicons and between amplicon lengths. (C) A heat map of correlation between triplicate amplicons and across the different sized amplicons identifying strong overall correlation (0.972 ± 0.024), but reduced correlation with the longer amplicons. (D) Box plots of the standard deviation of CpG site methylation levels between triplicate amplicons for all tested regions (low methylation: blue; intermediate methylation: grey; high methylation: red). Of note, moderately increased variability in CpG site methylation levels was observed with the intermediately methylated CpG island (MEST) and with amplicons greater than ~1 kb (see Results and discussion). Overlaid is a green line graph of sequencing depth for each amplicon, suggesting preferential sequencing of the shorter amplicons compared to the longer amplicons and the influence of reduced sequencing depth on CpG site methylation level reproducibility.