With all the hype around Batman-The Dark Knight, we missed one of the most recent heroic efforts to facilitate global methylation studies. The unveiling of a Bayesian tool for methylation analysis (Batman) enables researchers to estimate absolute methylation levels across a range of CpG dense regions using methylated DNA immunoprecipitation (MeDIP) approaches combined with either microarrays (MeDIP-Chip) or deep sequencing (MeDIP-Seq).
After its introduction in 2005 (Weber, M. (2005) Nature Genetics, Vol. 37, Number 8 ) the MeDIP approach has gained traction as an effective means to profile global methylation patterns on microarrays, and more recently, sequencing. Despite being useful, MeDIP hasn’t enabled researchers to quantify absolute methylation levels. Also, since the antibody used in MeDIP experiments is targeted at methylated cytosines, the density of CpG sites within a region can impact the overall efficiency of the immunoprecipitation, which can introduce additional headaches when analyzing data.
A team comprised of various UK institutions recently published (Nature Biotechnology, July 2008.) their approach that uses an algorithm that models the impact of varying CpG density on the enrichment process, ultimately enabling accurate measure of absolute methylation at a given locus.
MeDIP Captured
No matter what you are surveying, high density microarrays have revolutionized profiling and discovery experiments, but more recently they have become an invaluable tool to selectively subtract, or capture, regions of interest for downstream sequencing. In complex genomes like human, this approach as alleviated a major bottleneck in analysis and experimental cost.
Last fall, multiple groups published on the utility of sequence capture, but we have yet seen it applied to epigenetics until now (that doesnt mean it has not been done, we just havent seen it…hey we only have so many sets of eyes).
This UK group performed sequence capture on the MeDIP samples that enabled the enrichment of about 82% of known transcriptional start sites, 72% of non-promoter CpG islands, and other interesting areas scattered throughout exonic, intronic, and intergenic regions.
Batman Boosts MeDIP
Using Batman, the group was able estimate absolute methylation levels over a range of CpG dense regions, including CpG sparse regions, using both high density microarrays and deep sequencing. The data correlated well with bisulfite sequencing of randomly chosen regions. For all the details check out: Nature Biotechnology, July 2008