Solution decoder for protein folding.

[dlasecki]

Having constructed a full Hamiltonian for the problem, we compress the number of qubits by getting rid of trivial registers (initially reserved due to the size of a problem but then not relevant for optimization). Currently, this compression is 1-way only as we do not propagate information on which qubits were removed. It makes it impossible to interpret the solution vector - we are only left with an optimal value.

The goal of this issue is to propagate the relevant information and implement a method for reconstructing the full solution vector that is easy to interpret.