Dear Wandré,

You could try a larger energy range (i.e., 9) to generate more poses. Depending on your computer power, increasing exhaustiveness might be a bad idea. I suggest reducing it to 8-10 and increase energy range and num_modes.

P.s: I think Vina generates max 20 poses per ligand.

Definitely agree with Marcos, 

Though since you are talking about Rescoring take a look at --score_only and --local_only ( you can get information by typing $vina --help_advanced) and on the vina web page, 

more on that below.

Exhausiveness 400 is insane, the default is 8 and I've never seen a valid reason to go above the teens (10 to 16 generally puts you into more time for no gain already, maybe 20 if you are boxing a something huge ; but why would you want to do that)

See the Vina FAQ http://vina.scripps.edu/manual.html#faq "How big should the search space be"

OPENING up the range for --num_modes (9 as mentioned is the default for a reason) will generally give you several diverse posses, if not you can try increasing --energy_range to something >3 (in est kcal/mol)

That said if you are looking at consensus/redocking or rescoring:

Rarely will to programs give you Identical results, the question is what is close enough, and be very aware of how difference in set up,  or the nature of the program effect things.

if you are doing screening and the like: regardless of the methods, RANK-ORDER + cutoffs/thresholding is your friend

For Position based consensus scoring - Use an RMSD cutoff for similarity and be sure the box is properly positioned and big enough.

if you are screening for new interactions to test don't worry about just prioritize based on rank order similary (ex. both first 10% of the enery/score list for both methods, w/ vina kcal/mol

So, if I understand, the crystalographic pose can be the best pose, but, can be another rank or even not be predicted by vina. Is that right?

In my qualification exam, they mentioned that my poses predicted by vina can be better, but, any changes that I made, the results are not better than my previous results.

There is another docking software that I have to use or my parameters in Vina are not right?

Sorry for the wait,

The correct answer is 'it depends'.  -- (commentary on the crystalography verses ... at bottom)

how good is the crystal structure, is a  Holo, or Apo  in its structure,  is the ligand the same/similar/different?  (if the answer is Good, holo, same, or nearly identical try the rescoring method mentioned above - it will give you the values on the inputed position)

if the initial structure and parameters seem like they should work, the first thing I would suggest is to double check the the search box -- there is a common tendency to miss size, or miss locate it.

next look to alternate conformations and different rotimer combinations (possibly more) this goes in two directions, one the rotatable bonds in the ligand (and check the whole ligand while your at it, R,S etc, are correct and the given flexibility will allow for the empirical conformer)

if there is anything off, or ambiguous about the binding pocket you can also try the flexible side chain option (see the website), if access to the pocket is physically occluded this ?could? make it more accessible if for some reason it isn't being picked up by the initial search seeds.

On crystallography -- I'm not sure where this prevalent Idea comes from  (certainly no the crystallographers I talk to).

After hand waving away the modeling and computational effects and all the other fuzzyness. what you you end up with is a single data (think singular plural) representing the intersection of two basic things the 'solution' energy is close to a local minima, and without leaving the minima region, is able to form a symmetric cell -- you will often see deformation at the periodic boundary with allows for symmetry (with the deformation energy generally more than offset simply by becoming a solid, but occasionally helped by the evaporant )

the point being just from knowing the structure as a crystal you only 'know' 'that it is close to a energy minima so even a small amount in solution will spend a real time fraction in a similar conformation. This is great for speculating that something that is calculated to bind to that pose with bind at least as well in an experiment. But says nothing about best or optimal.

If you have a good grasp of thermo and know about the crystallization conditions (or better yet sets of trial conditions) you can squeeze more out of that -- but for that  talk to the crystallographers.