Are you using protein+ligand complex file in docking system as receptor? If yes then your results are obvious. If no then also that ligand may adopt another pose then crystal pose. Screen all poses and compare that crystal pose will have a slight positive value then your best docked pose (if all selected parameters and procedure is correct).

Dear Rajnikant,

I have removed the ligand beforehand before performing docking -- so my receptor is solely the protein. I have tried optimizing the exhaustiveness, but I received non-desired conformation.

Is there any other parameter that need to be optimized?

Dear Dr. Paul,

Thank you for a very helpful info. I really appreciate that.

I re-do the docking using Autodock4 but this time using the no-missing-atom complex PDB. I don't know how to ask for all binding poses, so I increased the number GA run to 50 -- it wrote 50 binding poses. I have checked all the poses, and I found one similar conformation with the RMSD 1.54 angstorm from the crystal reference, I am attaching the snapshot below. The cyan is from the crystal structure, and in pink is the docked conformation.

I would say my case fall under score error, since this conformation is in run 27th and alone in its cluster, not the lowest binding energy model. My question would be: Can this be considered as successful positive control docking? If I am going to perform virtual screening on the receptor, should I consider conformation from other cluster too?

I'm not sure what your level of expertise with Vina is, but here is a short tutorial I found on youtube that follows a very similar protocol that I use to prepare my ligands and proteins for docking. I would suggest not using the ligand from the Protein+ligand pdb file (even if you isolated it and cut it out and then added a flourine atom through Pymol - I have had the rare funny result when doing that before - and so now I draw all of my ligands separately, on Avogadroor ChemDraw).

https://www.youtube.com/watch?v=LlPSQBDNFr8

The only other times that I have had results that don't match well with experimental results is when using a large hydrophobic molecule - I'm not sure if that is the case with you.

I hope I was able to help, Good luck!

Hi,

what is your difference in scoring values between your lowest energy complex and your 27th one in your new experiment? Is it small? In this case, you'll be in one of the typical problem of scoring functions: there are not discriminative enough for your system.

best

@Shabir Ganai : I always have this doubt in my procedure. Do we need to minimize the protein complex beforehand? Wouldnt it make the pocket changes?

@Saul Wences: thanks! Ill take a good look on the tutorial.

@Dr. Maréchal: the lowest binding energy is -5.93 kcal/mol and my 27th is having -5.48 kcal/mol -- difference around ~0.5 kcal/mol. Any suggested cut off value for discriminative result?

@Dr Paul: Thanks for making it clear that both are actually done with different objecitves. :) But I didn't understand about the test you mentioned in the second paragraph. How can I score a pose without docking on it? :S

@Leandro Bortot: Thank you for making good highlights. :D No crystallographic water present in the interaction. Again, do we really need to minimize the protein complex?

I would like to thank all of you for sharing experience and knowledge. I really appreciate that.

Sometimes complex minimization is due. In the case of Vina and autodock this can be avoided since the programmes use unitede atom force field and you don't need to add all the hydrogens to the protein. But in case you use anothe program (i.e. plants, dock, surflex and so on) you need to make a short minimization to relax the added hydrogens. You cand do this with chimera.

Regarding the re.docking exercise you are doing, maybe the programs you use are not those you must use. Try chang docking program, there are a lot of free docking program.

If you have mukltiple complexes try to set up a cross-docking (ensemble docking) procedure to validate your program.

I am having problems with that too. You must consider the occupancy of crystal ligand, if the value is too low it is probably you won't find crystal position with docking methods.

I have a request. Thank you

https://www.researchgate.net/post/Is_any_one_with_expertise_in_molecular_docking_interested_to_join_a_collaborative_work_on_enzyme_inhibitors

Dear Zalikha Ibrahim , I am going through the same problem as you mentioned. Can you please tell me what did you finally do?

Dear Apoorva Kulkarni , last time I found out that the ligand was covalently bonded to the protein, hence I did slight modification on the method.

There are other things to be considered as well, for example, the size of docking grid box, the docking parameters etc.

Hope this helps!