Missing routing for what to do to actually make a fold, given that Grover's is really just a verification check for a good fold

[gravityshouldbenaut]

Grover's Search finds whether a specific characteristic exists within a set of qubits. Thus, it could theoretically be used for determining a fully folded structure, but not actually for performing a fold. To perform a fold, the easiest way to go would be to use QAOA to run to some sort of folding hamiltonian (Similar to what IBM did https://www.nature.com/articles/s41534-021-00368-4 or even what ProteinQure did https://arxiv.org/abs/1810.13411). But, this would lack any real chemical basis unless we encoded electronic or other states into our hamiltonian. What would be a good way to try this? And how many folding methods do we want to support? Topoly, for instance, can already do folds (https://pypi.org/project/topoly/) so we could just even use that and then use Grover for just the verification, but again the chemical basis is lacking.

For all these reasons, I think a modified VQE would be a good way to go, particularly this flavor adopted by Booth and the team at Rahko, wherein we decrease the total number of electronic states we are looking at and only look at high energy ones that really shape a structure (https://arxiv.org/abs/2104.05531). But, open to others.

[gravityshouldbenaut]

Not using Grover's Search anymore