To model the tethering experiments in silico the ligands should
covalently dock to the flexible side chain of cysteines (e.g., Cys195
of the active site of hTS 1hvy). A covalent docking is possible in GOLD
but it requires a linking atom existing in both, the protein and the
ligand. The linking atom should have an open valence electron (a
hydrogen atom bound to the linking atom of the ligand will be displayed
but not considered in the scoring function). The flexibility of a
protein residue can be specified in GOLD by adding manually a rotamer_lib
block to the gold.conf.
Protein and ligand have to contain the same linking atom with an
open valence electron.
A second sulfur atom (SG2)
was added to Cys195 using
pymol.
The linking atom of the ligands can be either specified by giving
the atom number or by doing a substructure search (substructure.mol2).
The substructure search works quite good for the tioles.
The flexibility of protein residues is only available for
GOLDScore.
gold.conf has to be changed manually and
not via the GOLD graphical user interface.
For each flexible main and side chain a rotamer_lib block must be given.
Problems
How to combine the flexibility of a protein residue with covalent
docking?
How to remove parts of the ligand?
R1-S-S-R2 -- R1-S-S*- and R2-S-S*- -- R1-S*-Cys and
R2-S*-Cys
How to replace parts of the ligand?
R1-S*-H -- R1-S*- and -S*-Cys -- R1-S*-Cys
Problem 1. In GOLD it is not
possible to use the flexibility of a side chain (not shown) to move the
linking atom. The linking atom of the ligand is everytime at the same
position.
(question
to the GOLD support)
A possibility to solve this problem is to use the Cys side chain as a
part of the ligand and attach it to the backbone CA atom of the protein.
Problem 2 and 3. So far
unsolved!
May be possible with ChemAxon's
Reactor
