Question on calcHairpin dg discrepancy

[sooheelee]

I'm observing a discrepancy in the dg given by calcHairpin for 'TTAACCTCAGATCCTAAGCCGCACAAAGTTGCGGCTTAGGATCTGA', a sequence from a paper I am reading.

• In blue is primer3's hairpin dg on y and primer3's hairpin Tm on x for four sequences including that above, which is the last datapoint.
• In orange is another software package's hairpin dg on y.
• The slight offset is due to minor differences in parameterization.
• Three of the sequences give concordant dgs for the two softwares

I expect the dgs to correlate with each other and to the hairpin Tm. Any idea why I see an aberration for this particular sequence?

Thank you for your time.

[sooheelee]

@untergasser ^

[bioinfo-ut]

I don't understand what are you trying to achieve with this analysis. Primer3 is NOT a software for calculation of hairpins and their strength. It's primary purpose is to design PCR primers that won't fail. Are you using this sequence as a PCR primer? Did it fail in an experiment?

If you want to calculate accurate secondary structures for DNA sequences of any length, then I would rather suggest MFOLD: http://unafold.rna.albany.edu/?q=mfold/DNA-Folding-Form

[ritah-nabunje]

@bioinfo-ut Is it okay using ntthal to obtain dG values for the formed secondary structures? Also, are the dG values in cal/mol or kcal/mol?