DEEPer begins with the selection of a set of perturbations to use (this step may be manual or automated). These perturbations define a set of RCs for each residue. RCs that cannot be in low-energy conformations are pruned, using the new indirect pruning algorithm (Section 2.4) as well as previous pruning algorithms. A* with minimization (see Introduction), as in iMinDEE [28], is then used to output low-energy conformations. The result is the GMEC and a gap-free list of all conformations and sequences within a user-specified interval Ew of the GMEC. This list can be used to select mutant sequences to synthesize for experimental testing, either by (i) selection of the sequences with the lowest-energy conformations, or (ii) by a provably-good approximation algorithm to calculate the binding affinities via the K* software module [5, 16].