How to recognize primer-dimers? - from a melting curve??... (Apr/23/2009 )
Hi again,
Iīm building up my real time PCR method and I have ran test runs and optimization runs lately. I have also ran dissociation curves after each run.
My questions are:
1) how do I know which peak is primer-dimer and which is the real amplicon? Obviously the amplicon has a higher melting point, but what about if there is only one peak? What is the melting point approximately to a primer-dimer (primers ~20bp and <50%GC)? First when I got two peaks in the melting curve, first at 69C (or not actually a peak but the curve went down from the top and the melting point was set at 69C in these samples) and second at 81C, I was sure that the first one was primer-dimer and second one amplicon (135bp, 49%GC).
But, when I ran optimization run for primer concentration, I got in some wells one peak at 81C, in some wells two peaks side by side at 81C and 83C and in some three peaks with third one at 85C. What are these later peaks with higher melting point? What could cause them? I started to doubt that maybe the actual product is the one with higher melting point...
Second:
2) I use relative quantification (not in optimization but in experiment) and I canīt add a dissociation stage straight to the end of the run, but I have to run a separate dissociation run. Does it matter if after the actual run there goes some time before I start the dissociation run? I mean does there happen something in the wells as the plate obviously cools down after the run.
Thanks alot again!
rajah on Apr 23 2009, 07:04 AM said:
Iīm building up my real time PCR method and I have ran test runs and optimization runs lately. I have also ran dissociation curves after each run.
My questions are:
1) how do I know which peak is primer-dimer and which is the real amplicon? Obviously the amplicon has a higher melting point, but what about if there is only one peak? What is the melting point approximately to a primer-dimer (primers ~20bp and <50%GC)? First when I got two peaks in the melting curve, first at 69C (or not actually a peak but the curve went down from the top and the melting point was set at 69C in these samples) and second at 81C, I was sure that the first one was primer-dimer and second one amplicon (135bp, 49%GC).
But, when I ran optimization run for primer concentration, I got in some wells one peak at 81C, in some wells two peaks side by side at 81C and 83C and in some three peaks with third one at 85C. What are these later peaks with higher melting point? What could cause them? I started to doubt that maybe the actual product is the one with higher melting point...
Second:
2) I use relative quantification (not in optimization but in experiment) and I canīt add a dissociation stage straight to the end of the run, but I have to run a separate dissociation run. Does it matter if after the actual run there goes some time before I start the dissociation run? I mean does there happen something in the wells as the plate obviously cools down after the run.
Thanks alot again!
The fact that you are getting multiple peaks leads me to believe that you have more than one product. A few questions: Were the wells with multiple peaks higher or lower primer concentrations? Have you BLASTed your primers and sequences to make sure you should only get one product? Have you cloned your product for sequencing? When in doubt, try running your real-time product out on a gel and see if you get multiple bands.
For the first question :
From what i've seen, primer dimers would indeed have a lower melting point than your amplicon. As for the theoritical melting point of a 50% GC-20 bp primer dimer.. I don't have a clue ^^. The fact that you have multiple peaks (69, 81, 83 and 85) could mean that your primers are not that specific, or that you need to increase the Tm of some degrees. It seems to me that those are non-specific amplicons.
Second question :
My guess would be that the shorter you wait, the best it is. I would also keep the plate on ice, to prevent any other amplification.
Good luck!
gfischer on Apr 23 2009, 05:33 PM said:
rajah on Apr 23 2009, 07:04 AM said:
Iīm building up my real time PCR method and I have ran test runs and optimization runs lately. I have also ran dissociation curves after each run.
My questions are:
1) how do I know which peak is primer-dimer and which is the real amplicon? Obviously the amplicon has a higher melting point, but what about if there is only one peak? What is the melting point approximately to a primer-dimer (primers ~20bp and <50%GC)? First when I got two peaks in the melting curve, first at 69C (or not actually a peak but the curve went down from the top and the melting point was set at 69C in these samples) and second at 81C, I was sure that the first one was primer-dimer and second one amplicon (135bp, 49%GC).
But, when I ran optimization run for primer concentration, I got in some wells one peak at 81C, in some wells two peaks side by side at 81C and 83C and in some three peaks with third one at 85C. What are these later peaks with higher melting point? What could cause them? I started to doubt that maybe the actual product is the one with higher melting point...
Second:
2) I use relative quantification (not in optimization but in experiment) and I canīt add a dissociation stage straight to the end of the run, but I have to run a separate dissociation run. Does it matter if after the actual run there goes some time before I start the dissociation run? I mean does there happen something in the wells as the plate obviously cools down after the run.
Thanks alot again!
The fact that you are getting multiple peaks leads me to believe that you have more than one product. A few questions: Were the wells with multiple peaks higher or lower primer concentrations? Have you BLASTed your primers and sequences to make sure you should only get one product? Have you cloned your product for sequencing? When in doubt, try running your real-time product out on a gel and see if you get multiple bands.
Thanks for the replies! qfisher:
The wells with multiple peaks were mainly the ones with higher primer concentrations for the target gene (BDNF) and the ones with lower concentrations for housekeeping gene (GAPDH) (!?!) which also is weird for me... The primers are BLASTed and used by many others as well as our collaborator so they should be good. I havenīt cloned the product for sequensing. Iīve run (normal)PCR products done with these primers (before real time to check possible contamination) on a gel and I get a good BDNF band and always some primer-dimers, no matter what I do (Iīve tried to change concentration/annealing temperature...)... maybe thereīs something wrong with my primers, maybe Iīll order new ones. Also what led me to this conclusion is that once I tried to test the efficiency of BDNF and GAPDH reactions I got nice standard curves for GAPDH but very weird ones for BDNF... argh, getting frusturated
when you set up your efficiency assays, it is best to do a double-titration of primer and template. even if you have chosen very good primers, the wrong concentration can result in dimers, non-specific binding, etc. adding too much is WAY worse than adding too little, but performing the double-titration will tell you how much is just right and will give you optimal efficiency. if you're getting dimers (extra peaks in the melting curve) then none of your data is good.
good luck!