1) To check for structural features in a protein, one should analyze the ramachandran plot using procheck. Besides this some additional constraints can be further studied using SAVES server that gives scores such as WHATIF, PROVE, ERRAT, VERIFY_3D.

2) I agree to the second answer by Gnanavel and 3) while using protein structure, it is advised to consider X-ray diffraction method if choice is there between x-ray diffraction and NMR. Reason is same as resolution and proper structural details are available by X-ray diffraction method. For docking method, resolution of protein structure is important and this structure should be complete. if there is any missing residues, then minimization and addition of residues is really important . This process is completed first before docking. Check for missing residues, valency and stability of conformation through in silico method. There are various commercial softwares such as Discovery studio, Schrodinger etc. as well as freely available such as Swisspdb viewer etc.

I do not have any experience with Hex.

But the docking experiment always can not give correct value.  People use just to compare the energy values between different legands.

One can get the correct result after a small dynamics and minimization only.

I hope it will help.

As Gnanavel said, modeled structures need to be minimized. Even some controlled MD in appropriate surrounding (water, lipid bi-layer...) is recommended for them.

Otherwise, avoid heavy minimization on crystal structures. If you think that minimization is necessary (e.g. low resolution structure), you can put some constrain at least on backbone atoms.

If you have good crystal structure (high resolution), usually you will avoid minimization. You could, after you check structure for other problems (missing or incomplete residues) and add hydrogen atoms, run minimization with heavy atoms constrained, so that only hydrogen atoms are minimized. That would clear structure of possible clashes.

You may check the energy difference before and after minimization. The difference should be in the form of +ve, then yours results are supposed to be correct. One thing more it is compulsory to do minimization of raw protein for the docking study.

Protein-protein docking is more complicated than protein-ligand due to a higher number of freedom degrees in side chains of both macromolecules involved (and also in main chains). I would suggest first to look for homologous complexes to be used as scaffold for simultaneous modeling of both chains, and hence of the complex. If this is not feasible, it would be interesting to send to hex (or equivalent) different conformers of the modeled monomers looking for alternative docking positions. These conformers might be the minimized and original versions of the monomers. A good procedure would be to check the structural quality of the complexes before and after energy minimization of the hex results. In my opinion the best structures would be those of higher stereochemical quality than those of the lowest energy. 

Good luck!

PDB structures does not have missing residues which are important for docking. @ Amandeep Kahlon please clear this doubt, as m bit confused.

Generally, i always take structure from PDB or model using any of the modeling approach and EM using UCSF chimera, SPDBV.

Steps are common

1. Add polar Hydrogen atoms

2. Add charges

3. Energy minimize the structure.

validate the refined structure and earlier modeled structure using RAMPAGE server- which gives Ramchandran plot. 

When we can minimize the protein? what are the criteria we have to look before doing minimization?

its better to use a structure with lowest energy when you want to perform docking, but it depends on the kind of software which you would like to applied for protein-protein docking, for example, if you use HADDOCK server, you don't need to do energy minimization, because one of the parts of this server does energy minimization.

if you perform energy minimization by molecular dynamics, you can find  the best structure in terms of the lowest  energy by determination RMSD and RMSF and ramachandran plot. in order to do energy minimization, you can use several software, like hyperchem, spdv, gromacs, . . .

3. In such case can we use normal protein (without minimization) for docking?

if you use a structure which was modelled by homology modeling, you should do energy minimization before docking.

best wishes

Marzieh

Crystallographic structures may have problems with missing residues, added waters, chain breaks etc. First of all these problems are corrected which require minimization of the protein. After then consider for docking otherwise results will not be correct even if the energy value being lower.

Hi

receptor minimization is not necessary always. if the structure you use is a complex(for example, a ligang is crystalled with that) the minimization is essential because the conformation of the protein is not in the native form, if not,  minimization is not necessary.

with reference to the question asked by Lalit, I would like to refer to PDB structure file when opened in wordpad to check for atomic details, residues, chains, etc.

I am giving one example of PDB ID: 1T2P (sortase A from Staphylococcus aureus). Please open this file you will find all the details of structure including method, authors, structural details. During the X-ray diffraction and NMR, some residues, atoms are located in the crystal structure diffraction data. These conditions effect the protein stability. Hence, energy minimization of protein is important before docking. 

Example: 1T2P

REMARK 465 MISSING RESIDUES

REMARK 465 THE FOLLOWING RESIDUES WERE NOT LOCATED IN THE

REMARK 465 EXPERIMENT. (M=MODEL NUMBER; RES=RESIDUE NAME; C=CHAIN

REMARK 465 IDENTIFIER; SSSEQ=SEQUENCE NUMBER; I=INSERTION CODE.)

REMARK 465

REMARK 465 M RES C SSSEQI

REMARK 465 ALA A 61

REMARK 465 VAL B 168

REMARK 465 LEU B 169

REMARK 465 ASP B 170

REMARK 465 GLU B 171

REMARK 465 GLN B 172

REMARK 465 LYS B 173

REMARK 465 GLY B 174

REMARK 465 LYS B 175

REMARK 465 ALA C 61

REMARK 470

REMARK 470 MISSING ATOM

REMARK 470 THE FOLLOWING RESIDUES HAVE MISSING ATOMS(M=MODEL NUMBER;

REMARK 470 RES=RESIDUE NAME; C=CHAIN IDENTIFIER; SSEQ=SEQUENCE NUMBER;

REMARK 470 I=INSERTION CODE):

REMARK 470 M RES CSSEQI ATOMS

REMARK 470 LYS B 62 CG CD CE NZ

REMARK 470 PRO B 63 CG CD

REMARK 470 LYS C 62 CG CD CE NZ

REMARK 500

Hi...

My suggestion is to go for energy minimization for even pdb crystal structure after corrections done in the pdb file, because the crystals formation was at some temperature so any protein get its crystallised form with certain respected local minimum conditions so better do energy minimization to get the that local minimum to at least near global minimum (also one of local minimum with less energy) to get behaviour of reasonable protein properties. Then proceed with usual molecular docking and dynamics procedure.

Thank you.