Sign up ×
Biology Stack Exchange is a question and answer site for biology researchers, academics, and students. It's 100% free, no registration required.

I am trying to predict the secondary structure of certain predicted pre-miRNAs through mFOLD which is the generally accepted technique for structure prediction in most studies. I am finding it hard to interpret the result and accept if the pre-miRNA which I have picked and the miRNA I have can be accepted and included in my result.

This is one predicted pre-miRNA structure with the predicted miRNA position shown in green. enter image description here

Is this a good/bad structure and if so, can anyone give me a reason why. Thank you.

edit: I understand that energy levels are also something that is looked for in the prediction. In this particular figure, ΔG = -75.00 kcal/mol which I think is fine.

Sequence used for folding is


and the highlighted match area is


So this structure should be fine because the sequence lies along a stem-loop structure. enter image description here

I am worried about that bulge on the left top corner though. That doesn't affect the predictability of this structure does it?

share|improve this question
As a non-specialist reader, I wonder when you say mMFOLD is a "technique" if you mean it's software? –  daniel May 26 '14 at 14:44
@daniel yup software –  The Last Word May 26 '14 at 20:30
Can I know more about the settings of this prediction? Temperature, ionic strenght, molarity etc? You could post the link to the job results on the mfold server so I can better evaluate the likelihood of this structure. –  alec_djinn Jun 19 at 10:58
@alec_djinn I have moved past this project but I generally stick to default parameters for prediction. The paper titled "Bioinformatic discovery of microRNA precursors from human ESTs and introns" gave me a guideline for identifying a viable precursor which I follow currently. But it would of course be helpful to know how the prediction is done. I will make sure to contact you the next time I do a precursor prediction and give you a link. Thank you. –  The Last Word Jun 20 at 4:35

1 Answer 1

up vote 1 down vote accepted

The green region will definitely not make an miRNA: It is not a part of a stem loop. See typical miRNA structures from miRbase.


Bulges can sometimes determine which strand is chosen as mature miRNA. However, this green region is also is unlikely to form miRNA because the stem is just 15bp.

share|improve this answer
Please look at the edit of the question. –  The Last Word May 27 '14 at 10:20
I didn't know if I should ask this as a separate question so, I am posting it as a comment. In one paper (, I found that ~70 nt of the matched flanking area was taken as pre-miRNA while in this paper ( ~100 flanking nt of the match area was taken for pre-miRNA structure prediction. Any idea why so? –  The Last Word May 31 '14 at 7:29
no idea.. it is best to take 60-70 nt from either side (because you don't yet know that your read is -3p or -5p). I suggest that you use a machine learning algorithm to predict the right miRNA structure. Co-incidentally a friend of mine was working on similar problem and was talking to me about it (for a moment I thought you and her are same people :P ). I'll let you know if we make some progress. –  WYSIWYG May 31 '14 at 8:11
I'll post it as a question then to see if anybody else could get some light on the issue.. –  The Last Word May 31 '14 at 9:04

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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