I know that PacBio-based sequencing can be used to identify methylation without special library preparation.

Besides this, pyrosequencing should principally be usable for detection of methylation if the sequencing library has been prepared accordingly, s.a. http://en.wikipedia.org/wiki/Bisulfite_sequencing.

Suitable, yes. See Nature Protocols 2007 "DNA methylation analysis by pyrosequencing" [http://www.nature.com/nprot/journal/v2/n9/full/nprot.2007.314.html]. However, the methodology you choose should be determined by the question you are asking in terms of the DNA regions from which you want to recover methylation information, the degree of sample preparation and manipulation you want to do, and the complexity of the downstream analytical bioinformatic processing you plan to pursue.

We do a lot of pyrosequencing, and it by far and away the gold standard for detecting DNA methylation at the single CpG / single gene promoter level.

There are several targeted techniques available, bisulphite-Sanger sequencing; bisulphite-pyrosequencing; and sequenom / Epityper mass spectrometry. All work on amplified DNA from your target region. Bisulphite/Sanger sequencing is almost never used because of the additional cloning step prior to sequencing, and the mass spectrometry techniques have the disadvantage that you often can't say exactly which CpG position in the amplicon was methylated. Bisulphite pyrosequencing is fast and reliable, however, the assays can take some time to set-up.

If however, you want a more global approach, then the Infinium arrays are the best technique, however they only cover a small percentage of the CpGs (but that is another story).

With the price of whole methylome (or Infinium style targetted CpGs throughout the whole methylome) being so low, a lot of people are doing this, rather than targetted sequencing. Bisulphite pyrosequencing is then the technique of choice for validation.

Hope the answer is useful

Jon

Yes, for sure. Mathias Ehrich from Sequenom has quite a publication record in this area. http://scholar.google.com.au/citations?user=A9UgSNEAAAAJ&hl=en

@Cedric Laczny, you need to convert 5mC to 5caC with Tet1 to call methylation using PacBio, the old circular SMRTbell library prep doesn't cut it. Also, with the PacBio machines low throughput, it's really only useful for bacterial epigenetics (or maybe fungal epigenetics), way too low for mammalian, or other large genome, methylation.

In the mammalian context, PacBio reads are best used for scaffolding, detecting structural rearrangements or reading through difficult to sequence low complexity repeats.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3598637/