Skip to main content
Fig. 1 | Genome Biology

Fig. 1

From: Genetic impacts on DNA methylation: research findings and future perspectives

Fig. 1

A typical workflow for meQTL identification. Step 1 is DNA methylation profiling. The most commonly applied methylation profiling technologies in meQTL studies are Illumina methylation arrays and whole genome bisulfite sequencing (WGBS). In both approaches, DNA is treated with bisulfite, converting unmethylated cytosines into uracils, and leaving methylated cytosines unchanged. DNA can then be profiled by sequencing or by Illumina array technologies, consisting of pre-designed probes. In step 2, DNA methylation levels at each CpG site are quantified, typically either as percentage (0–100%, e.g., in WGBS) or proportion methylation (0–1, e.g., in the Illumina technology methylation β-value). The example shows the distribution of methylation β-values for one CpG site (m1) across all profiled samples. Step 3 is the association of a set of genetic variants (coded as allele dosages at each locus) with methylation values at each CpG site, usually using linear models. In this example, after the association test at site m1 with a set of i genetic variants (shown in the Manhattan plot), g1 was found to be significantly associated with m1 (shown in the boxplot). Finally, step 4 represents the extension of the genetic association test to all profiled CpG sites genome-wide and the identification of genome-wide meQTLs after setting an appropriate threshold for statistical significance. The resulting meQTL associations can be either short-range, in cis (shown in heatmap for a few Mbp), or long-range or on different chromosomes, in trans (shown in Circos plot with all chromosomes)

Back to article page