Check complex base pair probability in relation to other strands not part of the complex

[UnHumbleBen]

Currently, the ComplexConstraint we are working on only computes base pair probability for a complex in isolation, but perhaps it should be calculated in relation to other strands in the design.

Another thing I am unsure about is that the NUPACK documentation actually suggest analyzing the complex with consideration of other complexes that could form

https://piercelab-caltech.github.io/nupack-docs/analysis/

As it is typically infeasible to experimentally study a single complex in isolation, we recommend analyzing nucleic acid strands in a test tube ensemble that contains the complex of interest as well as other competing complexes that might form in solution.

So in addition to considering other strands in the design, we might want to consider other complexes that could form.

[dave-doty]

One particular example would be to test that a single strand does not form too much secondary structure with a single-stranded overhang of a complex, for example. This could possibly be expressed as a complex constraint, where the complex contains all the strands of the original complex, plus the new strand, and asks that the new strand has low base pair probabilities for all its bases. But since NUPACK only looks at a particular circular permutation of strands, one might want to check many permutations so that secondary structure that can form in one permutation but not another is discovered.