AUTOMATED DOCKING USING A LAMARCKIAN GENETIC ALGORITHM AND AN EMPIRICAL BINDING FREE-ENERGY FUNCTION

A novel and robust automated docking method that predicts the bound co nformations of flexible Ligands to macromolecular targets has been dev eloped and tested, in combination with a new scoring function that est imates the free energy change upon binding. Interestingly, this method applies a Lamarckian model of genetics, in which environmental adapta tions of an individual's phenotype are reverse transcribed into its ge notype and become heritable traits (sic). We consider three search met hods, Monte Carlo simulated annealing, a traditional genetic algorithm , and the Lamarckian genetic algorithm, and compare their performance in dockings of seven protein-ligand test systems having known three-di mensional structure. We show that both the traditional and Lamarckian genetic algorithms can handle ligands with more degrees of freedom tha n the simulated annealing method used in earlier versions of AUTODOCK, and that the Lamarckian genetic algorithm is the most efficient, reli able, and successful of the three. The empirical free energy function was calibrated using a set of 30 structurally known protein-ligand com plexes with experimentally determined binding constants. Linear regres sion analysis of the observed binding constants in terms of a wide var iety of structure-derived molecular properties was performed. The fina l model had a residual standard error of 9.11 kJ mol(-1) (2.177 kcal m ol(-1)) and was chosen as the new energy function. The new search meth ods and empirical free energy function are available in AUTODOCK, vers ion 3.0. (C) 1998 John Wiley & Sons, Inc.