I am attempting to reproduce results from a number of journal articles all referring to the same SNP. In doing this I'm using the same primer set outlined in the articles. When I attempted a run the other day I experienced what I can only assume was primer dimer (~40bp product, bands not visible at expected bps.) While the primers and restriction enzyme were the same among numerous articles, the annealing temperatures in the reaction varied a fair amount (52-59C). All were above the individual Tms of the primers which struck me as odd.

Calculating the Tm via primer blast here are the numbers:

Forward Tm: 50.14 [5'-GCTCTACTTCCTGAAGACCT-3']

Reverse Tm: 51.67 [5'-AGTCTCACTCACCTTTGCAG-3']

Other sites gave slightly higher numbers (54-55) but they are still below the 57C that I used in my reaction. I do not have access to a gradient thermocycler otherwise I would have tried a range myself.

If an annealing temperature is chosen above the individual Tms how do the oligos have a chance to anneal with the template and eventually extend? I remember hearing the Tm calculation is more involved and includes finding where the melting curve changes direction. But on the surfance I would think the primers would never have a chance to anneal at such high temperature? What is wrong about my interpretation?

  • $\begingroup$ What kind of polymerase are you using? $\endgroup$ Commented Feb 1, 2012 at 0:10
  • $\begingroup$ Using a Maxime PreMix which includes Intron's i-Taq $\endgroup$
    – bobtheowl2
    Commented Feb 1, 2012 at 2:34

2 Answers 2


I don't think primer dimers are your primary concern here. Usually in my experiences, I get primer dimers all the time, even if the reaction works and I get my bands of interest.

Maybe you ought to troubleshoot other aspects of your PCR that might account for why your reaction isn't working. Have you tried using a positive control with your primers? You may try varying the parameters of your PCR as well. Remember, standard Tms are calculated according to a 50mM salt concentration: is that what you're using?

Generally, if the annealing temperature is above the predicted Tms of your primers, this represents a more restrictive and selective amplification of your target. You usually use a high annealing temperature if you're seeing lots of non-specific products. Since you're seeing no products at all, consider lowering the Tm to that of your primers (50º).

If you start seeing non-specific products at that Tm, I'd do what we call a "Touchdown" PCR -- that is, you start the reaction at a higher annealing temperature, and as each cycle progresses, it "touches down" to a lower annealing temperature. The principle behind this is that it starts it off at a restrictive temperature -- so your yield is very low initially, but then by gradually decreasing the "restrictiveness" of the reaction, your yield will improve. This will still prevent non-specific amplification b/c the less restrictive amplification will be on the fragments already amplified from the restrictive condition. Remember, biology is not exact science. Just because something says the Tm is such and such does not mean it's absolute and doesn't give you some margin of error. For example, we exploit that margin of error to optimize our experiments in the example I outlined above.

Anyways, (might have went on a tangent a little bit) you also have to consider how you designed your primers. Primers wil high self-complementarity will self-anneal at higher Tms.

  • $\begingroup$ "biology is not an exact science" made me chuckle - just don't tell the chemists and physicists or they'll never let us live it down! $\endgroup$ Commented Feb 1, 2012 at 0:36
  • $\begingroup$ @RichardSmith They way I learned it: Me: So, where can I find a volumetric flask? Post-doc: scoffs Who do you think we are? Chemists? :) $\endgroup$
    – jp89
    Commented Feb 1, 2012 at 5:43
  • $\begingroup$ Ironically this is one of the few times were I didn't design the primers and used what was mentioned in numerous papers. $\endgroup$
    – bobtheowl2
    Commented Feb 2, 2012 at 19:01
  • $\begingroup$ @bobtheowl2 Hm.. This may be a last resort thing (and may take a million years) but you can always send an e-mail to the authors of those papers and see if they can help you optimize your reactions or share with you how exactly they did it. $\endgroup$
    – jp89
    Commented Feb 2, 2012 at 19:05

Primer Tm calculations can vary significantly based on the method used. What I can tell you, is that the Tm really depend on the polymerase you are using for the PCR reaction and for each polymerase there is a set of PCR conditions you have to follow. As I use Phusion® Hot Start High-Fidelity DNA Polymerase from Finnzymes, I will give you an example with it:

First of all, it is suggested to use the Finzymes Tm calculator to determine the Tms of your primers. Usually the Tms are higher than the ones calculated by IDT. Then, according to the length of your primers, different annealing temperature should be set:

  • for primers > 20nt, anneal for 10 – 30 seconds at a Tm +3°C of the lower Tm primer
  • for primers ≤ 20nt, use an annealing temperature equal to the Tm of the lower Tm primer

When calculating the Tm of your primers with the calculator: Fwd primer Tm is 57.27 Rev primer Tm is 60.21 As they are 20bases long, your annealing temperature should be 57 degrees.

You are right that the annealing temperature shouldn't be higher than the Tm of the primers. I think the problem is that different polymerases and different Tm calculators are used in the papers you are referring to. I can discuss other PCR optimizations but let me know first whether the problem is not just the anneaning temperature.

  • $\begingroup$ I like your example and calculation. I guess I didn't think Tm was so polymerase specific before. Coincidentally it ends up at the same annealing temperature that I tried though. Unfortunately the taq I am using gives very little information (and not an online calculator) in regards to choosing temperatures. I am close to ordering a new mastermix anyway, so I will likely try a different brand then. But as for my current problem there are enough variables that I plan to try the same annealing temperature again and if that fails completely then I may look at what @jp89 suggested also. $\endgroup$
    – bobtheowl2
    Commented Feb 2, 2012 at 19:00
  • $\begingroup$ You can also contact the authors of some the papers you are trying to reproduce and ask them about the specific conditions they used. You shouldn't spend much time troubleshooting this by yourself. $\endgroup$ Commented Feb 2, 2012 at 19:10
  • $\begingroup$ Hey what happened with your experiment? $\endgroup$ Commented Mar 3, 2012 at 0:47
  • $\begingroup$ You can always use a lower Tm, but then you sacrifice specificity. In situations like these, I like using a gradient thermocycler to test a range of annealing temperatures to find the optimum. $\endgroup$
    – user560
    Commented Oct 8, 2012 at 1:44

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .