Motivation: Matching of chemical interacting groups is a common concept for
docking and fragment placement algorithms in computer-aided drug design. T
hese algorithms have been proven to be reliable and fast if at least a cert
ain number of hydrogen bonds or salt bridges occur: However, the algorithms
typically run into problems if hydrophobic fragments or ligands should be
placed In order to dock hydrophobic fragments without significant loss of c
omputational efficiency we have extended the interaction model and placemen
t algorithms in our docking tool FlexX. The concept of multi-level interact
ions is introduced into the algorithms for automatic selection and placemen
t of base fragments.
Results: With the multi-level interaction model and the corresponding algor
ithmic extensions, we were able to improve the overall performance of FlexX
significantly. We tested the approach with a set of 200 protein-ligand com
plexes taken from the Brookhaven Protein Data Bank (PDB). The number of tes
t cases which can be docked within 1.5 Angstrom RMSD from the crystal struc
ture can be increased from 58 to 64%. The performance gain is paid for by a
n increase in computation time from 73 to 91 s on average per protein-ligan
d complex.