Mysterious peaks in pyrosequencing runs!? - (Aug/06/2012 )
Our lab recently started to use pyrosequencing for quantitative methylation analysis. We all do not have a lot experience and expertise with this technique.
I am currently investigating the methylation status of one gene in tumor tissue samples. I have sequenced more than hundred samples and for a few samples I made the following observation: The pyrogram shows peaks at dispensations, where no peaks should occur. When I analyze the corresponding PCR products by gel electroforesis, I see a nice band at the good size, plus a slight band below the good band. It looks like a dimer, but why do I only see this in combination with the peaks? Could this also be contamination?
I always include a negative control (no DNA in PCR mix) in the pyrosequencing runs, which do not give any results (= clean). I checked my primers with MethylBlast and they should be specific.
I would be more than happy, if anyone could help me looking for a reason for those peaks/bands!! Did someone make similar observations?
I cannot analyze my data until I know, what I am dealing with...
Most of the time this type of problem is due to inspecificity, either from the PCR primers, either from the sequencing primers. The fact that you observe an aspecific amplicon in your gel probably indicates the first reason is causing the unexpected peaks. Is the size of the aspecific amplicon what you would expect from primer dimers? If not, you have need to redesign your primers.
Thanks ForEpi for your reply!! At a frst glance, the gel bands look like primer dimer. They do not look like bands more like a shadow. However, I observed that peaks at an A-dispensation are associated with 'dimers' directly beneath the PCR band, whereas peaks at a G-dispensation are associated with 'dimers' located lower than the "A-peak dimers". I think you are right and what I see is unspecific binding. However, when I repeated the PCR+pyrosequencing run, the peaks and bands are gone. Is it possible that I pick up something unspecific the first time and the next time it is gone?