Methods Comparison for Methylation Analysis - Which method is the best? (Apr/06/2007 )

Hi all,

lately, a lot of new methods for DNA methylation analysis are available, including Pyrosequencing, MALDI-TOF based analysis (Sequenom MassArray+Epityper) or high resolution melting curve analysis.
Are you aware of any independent papers comparing these methods? Would be interesting regarding efficacy, hands-on-time, prize (purchase + reagents) etc.

Any idea? Or experience?
Krümel

I have used both sequenom epityper and melting curve analysis (we also have a pyrosequencer, but we havent tried it out for methylation analysis although we have had the training) and traditional direct sequencing/cloning and sequencing - these are my impressions

Obviously melting curve analysis is quick and cheap. and you get what you pay for - almost uninterpretable results! I guess Im being a bit harsh, but with my experiences of it, I would take it as an indication of the methylation level only... and dont get me started about multiple peaks, and shoulder peaks etc...

sequenom epityper is a viable replacement for direct PCR product sequencing - except higher throughput (384 samples on each chip). From PCR product to data acquisition takes about 3 days, and a automated liquid handler is very very highly recommended, otherwise your hand is going to be very sore
It is not a replacement for clonal sequencing, as with any technique that does not clone out individual PCR products you are getting an average value over all molecules in your PCR product.
it is quite accurate and precise (we can confidently detect a 10% difference at a single site between samples - now you only need a biological question that needs such accuracy!). The thing that they dont tell you is that you dont get data for every CpG site in your PCR amplicon - because its mass spec, it is the mass of the fragments generated that separates the CpG sites, as opposed to their position in the sequence. so often there are CpG sites within fragments that are too small or too large to be analysed by the MALDI-TOF. commonly multiple CpG sites in different fragments have the same mass and "pile up" on top of each other and therefore give an average value... but from all the amplicons i have analysed with epityper i would say ~50% of the CpG sites are interpretable

as for publications, watch this space

Thanks, that was very informative... (and I'd really like to know what kind of lab you're working - do you have a never-ending budget?)
Any comments on Pyrosequencing, anybody?

Another option for high throughput highly quantitative analysis would be the 454 genome sequencer - although i assume its really expensive!
just today i got a new application note for how they used it for clonal sequencing of a p16 amplicon - they got the equivalent of 12000 clones for the majority of the amplicon!

New paper out on this topic

Genomic profiling of CpG methylation and allelic specificity using quantitative high-throughput mass spectrometry: critical evaluation and improvements.

Coolen MW, Statham AL, Gardiner-Garden M, Clark SJ.

Nucleic Acids Res. 2007 Sep 13;

Hey @ all,

I´m using Pyrosequencing for my Methylation Analysis. Unfortunatly we don´t have our own Pyrosequencer, so we have to
send our samples to a company which is perfoming the Sequencing Reaktion.

But realy like this technique because you get a quantitative Analysis of your region of interest.

I look at spezific Promotor Regions or CpG Islands.

The "only thing" i have to do, is Bisulfite Treatment and PCR , then I send the PCR to the company und get ready to publish date back.
If you compare Pyrosequencing with classical Bisulfite Sequencing i would always prefer Pyrosequencing.
It´s in the end cheaper und less Time intensive!

( I cannot say annything about high troughput Analysis)