Abstract:
A new program package for docking of a flexible ligands into the active sites of proteins as well as into other biological targets is discussed. The initial validation is also described. It is known that the docking procedure is a main tool in modern computer-aided drug design processes. In our program we use the genetic algorithm to optimize the ligand's inner torsion degrees of freedom and a rotational-translational position of ligand as a whole in a target active site. The optimization is guided by estimating the inner ligand energy and the energy of ligand's interaction with biological micromolecules. The calculations of these energies are performed in the framework of the MMFF94 force field model. Solvation-desolvation effects are considered via Generalized Born Approximation. The program also calculates the estimate of ligand-target binding free energy and performs the clusterization of solutions (ligand's poses) according to its geometries. The data from the initial validation show that our program can perform the successful positioning of ligands in protein's active sites as well as virtual screening for active ligands in database which contains active and inactive compounds.
Keywords:computer-aided drug design, docking, genetic algorithms, global optimization, MMFF94 force field, solvatation model, virtual screening.