Discrepancies between UNAFold softwares

Ask questions, report problems or request new features for the UNAFold software.
Post Reply
tylerv1
Posts: 3
Joined: Tue Feb 20, 2024 3:43 pm

Discrepancies between UNAFold softwares

Post by tylerv1 »

Hello,

I recently obtained the UNAFold license from Rensaeller and I noticed that there was occasionally discrepancies between this software and the IDT version found at https://www.idtdna.com/calc/analyzer. Notably, when I input this sequence: "AAAAAGAATTACACTTATCAGAAAATTCAAGTAATTC" at 25C and 150mM [Na+], I only get one structure for the licensed software (in the .fasta.ct file [not the .fasta_1.ct file]), but IDT displays 3 total structures. The first structure is the same, however I am trying to minimize non-target structures and was wondering if there has been any updates to the IDT version not found in the licensed software.

Thank you so much,
Tyler
nmarkham
Site Admin
Posts: 24
Joined: Wed Jun 29, 2022 5:23 pm

Re: Discrepancies between UNAFold softwares

Post by nmarkham »

I'm not familar with IDT's "OligoAnalyzer Tool". I didn't even know offhand that it uses UNAFold.

Assuming they are, I don't know precisely which version of UNAFold they use. It could very well not be UNAFold 4, which I assume is what you have, although I wouldn't expect any differences that are all that significant. It could also be that they used different options than you did. If IDT showed you multiple structures, I imagine they obtained them with the --mfold option; did you use that as well? If you didn't request suboptimal tracebacks in the first place, that could be an obvious and fairly uninteresting explanation. :D

If you're interested in replicating their results exactly, I suppose your best bet is to contact IDT and ask them exactly what version of UNAFold and what options they use.

Hope this helps!
tylerv1
Posts: 3
Joined: Tue Feb 20, 2024 3:43 pm

Re: Discrepancies between UNAFold softwares

Post by tylerv1 »

Thank you so much for your prompt response! After doing more testing with the mFold webserver: http://www.unafold.org/mfold/applicatio ... g-form.php, I found that you were right and that the IDT website seems to use ~50% suboptimality number. Is there anyway to adjust the suboptimality using UNAFold and not mfold? I ask because I have a windows computer and thus I'm unable to setup mfold on my machine.
nmarkham
Site Admin
Posts: 24
Joined: Wed Jun 29, 2022 5:23 pm

Re: Discrepancies between UNAFold softwares

Post by nmarkham »

In general, if you want suboptimal structures, you use the "--mfold" option. Merely specifying it engages "mfold mode", where hybrid-ss-min (or hybrid-min or hybrid-intra-min, for that matter) generates a sample of suboptimal structures as well as an optimal one. However, --mfold can optionally take arguments to change things like the suboptimality percent. In particular, --mfold=50 might be what IDT is doing. See http://www.unafold.org/Dinamelt/unafold ... ss-min.php for more details about --mfold.

You (or anyone else who's interested) might also try the "Quikfold" app on this site: http://www.unafold.org/Dinamelt/applica ... ckfold.php. It generates suboptimal foldings and allows you to vary the three optional arguments to --mfold to change what sort of sample you get.
tylerv1
Posts: 3
Joined: Tue Feb 20, 2024 3:43 pm

Re: Discrepancies between UNAFold softwares

Post by tylerv1 »

Thank you so much for your help, the --mfold option using hybrid-ss-min worked perfectly!
Post Reply