Interpreting DNA methylation data is tough enough when you get binary, yes/no results, but can be downright maddening when your CpG sites are giving you mixed signals. Researchers at UCSD have found that it’s been allele-specific methylation (ASM) and CpG-SNPs causing much of that “fuzzy methylation”.
The scientists tackled the thorny issue of CpGs with intermediate methylation levels, called fuzzy methylation, (which does not refer to DNA grooming habits , as we first thought) by scanning ASM across 16 cell lines with a potent combo of bisulfite padlock probes (BSPPs), sequencing, and linkage disequilibrium (LD) analysis adapted from SNP calling methods. Their unique approach paid off with some interesting findings.
- ASM was observed in 23-37% of SNPs, depending on the cell line, a big chunk of which was the result of SNPs at CpG sites, disrupting their methylation potential.
- Overall, 38-88% of ASM regions are caused solely by CpG-SNPs, revealing that genetic variation at CpG sites is critical for ASM.
- 30-48% of fuzzily methylated CpGs correlate with ASM and 14-36% are located in LD blocks (100+ bp regions with 10 or more CpGs associated with ASM) showing stochastic effects don’t account for all of the fuzzy methylation that is seen.
- They estimate that there are over 200,000 SNPs located on CpG sites, meaning there is a ton of real-estate available for potential variation.
The authors concluded that CpG-SNPs are likely an important class of cis-regulatory variants that alter CpG methylation sites and therefore link genetic variation with the epigenome.
Clear up your fuzzy methylation questions at Genome Research, April 2010.