I am building primers for LAMP which I know will not match all my target sequences perfectly.

To some extent, having primers with multiple variants at variable sites (degenerate primers) and longer sequences (tagged degenerate primers) will help. However, I can't just increase degeneracy (variability) forever because then I will have no amplification efficiency.

Therefore, especially at the 5' end which is less important from an amplification perspective, I will have some calculated mismatches. So I have this puzzle:

If those positions of interest (closer to the 5' end) are likely to be any base with about the same frequency, which is either a) the base or b) the pair of bases, which are likely to clash the least?

My instinctive reaction is that for a single base, a pyrimidine would fit better, because even if it doesn't match, it's smaller? And for a pair of bases, probably a purine and a pyrimidine? I would appreciate your thoughts!

  • 1
    $\begingroup$ I'm not sure this is the right place to ask something like this since it is about the wet-lab side of things and not bioinformatics. I'm not closing it since I am not 100% convinced, but I suspect it would be better on Biology instead. $\endgroup$
    – terdon
    Commented Mar 6, 2023 at 14:07

1 Answer 1


I would just use Primer(3)-Blast here: https://www.ncbi.nlm.nih.gov/tools/primer-blast/index.cgi?GROUP_TARGET=on

I would avoid the approach you outline because:

  1. Primer design algorithms incorporating degeneracy are very good;
  2. I'm not sure you're right (but could be wrong).

The algorithm behind Primer-Blast is Primer3, and is legendary in the primer design business and of course Blast.

I agree that strong conservation at 3' is key to the design, but based on homology rather than chemistry.

The thermodynamic calculation ... I would suggest that a mismatch with AT is going to less severe than GC. GC is a stronger bond, but I'd let the algorithm sort this out and prioritise using personal judgement:

  1. Conservation by homology at the 3' end.
  2. Conservation by homology for GC regions.

Ultimately, primers are cheap so ordering loads of algorithmic designs is usually the best approach.

Just to add an explanation on the subject of why is this bioinformatics?

One of the issues of primer selection is self-hybridisation within the primer or annealing between the primer pair. This is bad news for a primer - because it might not work.

Algorithm primer design programs are good at spotting and avoiding unwanted hybridisation. When degeneracy is involved this becomes complicated.

What Primer-Blast does is assess all combinations of degeneracy against issues such as self-hybridisation and primer-annealing. Humanely it is not possible to assess all those combinations and that is why an algorithm is needed.

  • 1
    $\begingroup$ Thank you! As helpful as Primer3 is, it is not exactly optimized for LAMP, which is what I'm toying with right now. The alternative, PrimerExplorer5 was insufficient on its own. But it's good advice. $\endgroup$
    – Laura
    Commented Mar 7, 2023 at 16:45
  • $\begingroup$ Many thanks @Laura and thanks for the update on PrimerExplore5, I didn't know about it. A good question all round. $\endgroup$
    – M__
    Commented Mar 7, 2023 at 17:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.