Molecular Docking Algorithm Based on Shape Complementarity Principles

Question:

I get an error message about my input from the server. What is wrong with my input?

Answer:

Possible Input Errors:

• PDB code: make sure you enter valid PDB code, i.e. the molecule with this code exists in PDB.
• Chain ID: make sure that the chains you specified exist in the PDB file.
• Uploaded file: should be in PDB format only!

Question:

I specify 'antibody-antigen' complex type and get an error message. What can be a problem?

Answer:

The antibody molecule should allways be specified as the 'receptor molecule'.

Question:

I specify the binding site for one/both molecules and get an error message. What is the problem?

Answer:

The binding site file should contain the following line for each potential binding site residue:
[residue index] [chain ID][newline]...
Note: This file must be consistent with the given PDB file. The chain ID should be the same as in PDB file.
If there is no chain ID in PDB file, do not specify chain ID in the binding site file.
An example file site.txt:
88 L
89 L
90 L
91 L
92 L
93 L
95 H
96 H
101 H
102 H
If there is no chain ID in the PDB file, the site file should look as follows:
88
89
90
91
92
93
95
96
101
102

Question:

I can't see the ligand molecule in the output PDB file. What is the problem?

Answer:

The files generated by the server are best viewed by Rasmol. Rasmol can be downloaded here .
Generally PatchDock server does not modify your input files, except for the coordinates of the ligand.
It joins the receptor and the ligand file with modified coordinates.
If your receptor file had an END record at the end, this may be a problem for some viewers, so the ligand will be ignored.

Question:

I don't know how to select the near native solution out of the solution list?

Answer:

Selection of near native docking solution is one of the most difficult and challenging problems in Structural Bioinformatics.
Nevertheless, docking can be used in combination with experimental data, i.e. select the solutions
that fit some available experimental information, like binding domain, catalytic residues, mutations, etc.
PatchDock allows you to include binding site information in the docking search stage, in order to obtain only relevant solutions.

Question:

How to apply the 3D transformation PatchDock outputs to the ligand molecule?

Answer:

PatchDock outputs a ligand transformation that transforms the ligand onto the receptor. The transformation is specified by 3 rotational and 3 translational parameters. The rotation in PatchDock is in XYZ notation. The rotation matrix from the 3 rotation angles (x,y,z) is:

real cx = cos(x) real cy = cos(y) real cz = cos(z)
real sx = sin(x) real sy = sin(y) real sz = sin(z)

    (cz*cy, -sy*sx*cz - sz*cx, -sy*cx*cz + sz*sx)
M = (sz*cy, -sy*sx*sz + cx*cz, -sy*cx*sz - sx*cz)
    (sy, cy*sx, cy*cx )

Now use the matrix M to apply the whole transformation (rotation and translation t) on a ligand coordinate v:

transformed_v = M*v + t