Following prior suggestions, I made “heparin” molecules as a single ligand. In one case, I have two 12mers with the only difference being 3O-sulfation on the third monosaccharide out of 12 in one case, and no 3O-sulfation in the other case. Based on some experimental evidence, we believe that the 3O should be interacting but don’t know where with the protein. Can I force this interaction while using the HADDOCK 2.4 server? The HADDOCK results I have so far don’t favor this 3O-sulfation group.
You would have to define manually a distance restraint and provide e.g. as unambiguous restraints in the server (third tab).
How did you currently setup your docking? Especially if you don’t know where the interaction is on the protein.
Did you use center-of-mass restraints?
To define the restraints, best would be to limit the atom selections on the protein to the potentially interacting atom types.
Would that be mainly positively charged residues?
Here is an example of such a restraint (assuming here that the sulphate group oxygen atoms would be named O1S, O2S and O3S), and that the protein is chain A and the ligand chain B:
assign (segid B and name O*S) (segid A and (resn ARG or resn LYS or resn HIS) ) 2.0 2.0 0.0
Adapt it to your needs.
Note that this distance restraint can be combined with the center-of-mass restraint.
Sorry - I don’t think I told the fully story the first time. We do know roughly where it binds on the protein but not which direction it would bind in. For the first time I did it, I set the reisdues on the protein side and then just used #1 on the small molecule side because I re-wrote the glycan as a single small molecule. I did try a blind dock using center of mass, and it bound in the same place as when I guided it with the active residues.
Thanks for the tip about how to put in the distance restraint. I will try it.