In my perspective, the reason we perform the step of minimization is to lower the degree of stearic clashes that might occur in amino-acids with improper orientations. Also, after performing this step protein restores it relaxed state after which it is more feasible to move forwards with docking and other computational studies.

However, I haven't used the software which you have mentioned. My experience is with Schrodinger. Still. it would be better to perform the step either way.

If you do re-docking to benchmark your docking program, starting from a known complex, simply removing the ligand and docking it again is easiest (cheating), docking to the minimized receptor after removing the ligand, docking to an independent unliganded receptor structure and docking to a homology model are increasingly difficult tasks, as the receptor is increasingly different from the conformation in the native complex.

If you are really interested in finding the correct interaction and are not starting from a known complex structure, you will not just minimize the receptor to get a single starting structure, but produce an ensemble of slightly different starting structures, e.g. by MD or Rosetta backrub.

You may want to look at the CAPRI website (https://www.capri-docking.org/resources/) and the publications referenced therein (https://www.capri-docking.org/publications/) to see which methods were most successful in predicting protein-protein interactions by blind docking

e..g

Article Modeling protein-protein and protein-peptide complexes: CAPR...

Respected Annemarie Honegger mam and Muhammad yousuf sir

Thanks a lot for such detailed information, I am really grateful.