I know how to compute the Nussinov score for a given RNA sequence, but I don't know whether any score I compute is biologically significant. How do I go about computing a baseline score using the Nussinov algorithm such that any score above this baseline, I know is biologically significant?
One way of computing a 'background' to compare a score to, is be to shuffle/permute your RNA sequence and recompute the score for each such 'randomized' sequence.
I assume this is for illustrational/educational purposes only and that you know about the efficient algorithms and implementations which are based on proper thermodynamics models, e.g. the Vienna RNA package.