Successive docking

HADDOCK
1

Hi to everyone, I would like to get a feedback on the procedure I’m doing to possibly identify the binding interface between two proteins.
I have a ligand (~300 amino acids) and a receptor (~150 amino acids).

  1. First, I try to dock the amino acids 1-100 of the ligand with the receptor, than the amino acids 101-200 of the ligand with the receptor, and finally the amino acids 201-300 with the receptor (from these run usually I get a high restrain energy ~1000/3000).
  2. For any run i collect the best cluster according to the haddock score
  3. I determine the list of amino acids in contact between the three “best” clusters and the ligand
  4. I perform a new docking constraining the contact between the list of amino acids of the ligand and receptor I got from the step 3 (from these run I got a lower restrain energy, but still high ~200/400)
  5. eventually re-determine the list of amino acids in contact from the best cluster and re-run the docking using the new list

My questions are:

  1. Does this scheme make sense according to the way haddock works?
  2. Are the final value of the restrain energy reasonable or still too high?
  3. which force field is used to compute the energy terms of the molecular complex? (I would run a gromacs simulation to check the stability of the “best” complex I get at the end)

Thanks for any help, and happy new year

2

Why cut the “ligand” in three pieces? Is it an disordered protein?

If not, rather perform some ab-initio docking with haddock and analyse the contacts made at the initial rigid body stage, as explained for example in our protein-ligand binding site tutorial:

Try also some bioinformatics binding site predictions. Plenty of servers for doing that including our CPORT and WHISCY servers.