Multi-conformational ligand representation in 4D-QSAR: Reducing the bias associated with ligand alignment

Quantitative structure-activity relationship (QSAR) is an area of computati onal research which builds mathematical or atomistic models to predict biol ogical activities of molecules. While more powerful approaches make use of a genetic algorithm to reduce the bias with respect to model construction, the predictive power of the resulting surrogate still critically depends on the spatial alignment of the ligand molecules used to construct it. The 4D -QSAR concept Quasar developed at our laboratory not only takes local induc ed fit and H-bond flip-flop into account but also allows for the representa tion of the ligand molecules by an ensemble of conformations and/or orienta tions. The contribution of a single entity within this ensemble to the tota l ligand-receptor interaction energy is determined by a Boltzmann criterion . The three-dimensional surrogate is represented by a family of receptor-su rface models, populated with atomistic properties-hydrogen bonds, salt brid ges, hydrophobic particles, and solvent-mapped onto it. Quasar has been used to establish QSARs for the enzyme dopamine beta-hydrox ylase and for the aryl hydrocarbon receptor. The surrogates were able to pr edict free energies of ligand binding, Delta G degrees, for external sets o f 15 and 26 test ligand molecules, respectively, to within 0.7 kcal/mol (rm s) of the experimental value, with the largest individual deviation not exc eeding 1.3 kcal/mol. The results indicate that the use of a multiple-ligand representation is superior to a single-conformer concept and reduces the u ser bias associated with the ligand alignment. Moreover, the selection prot ocol demonstrates that the technique Is capable of identifying a small numb er of active conformations and does not prefer a larger selection of lesser -contributing entities.