Thank you for your hint (here, and in the email). I tried mfold, and it
gave the same result as the Vienna package. In fact, when I asked for
suboptimal structures, the second-best was the structure with all 9
pairs. So, I supposed that the reason is a strain in the loop. But then,
I don't understand why RNAdraw (which uses the same algorithms) DID join
that base pair, and instead, failed to join the two bases at the other
extremity of the stack (where there is no loop). Different energy
parameters? May it be so sensible to that?
William Knight wrote:
> Lack of a-u base pair on the right is probably because it would cause
> strain in the loop. Trying folding with mfold and viewing some of the
> resulting secondary structures.
>> William Knight