Small-molecule hydration energy and binding affinity calculations using quantum mechanical approaches

AUCAR, MARÍA G.; ADLER, NATALIA S.; CAVASOTTO, CLAUDIO N.

Buenos Aires

Congreso; Frontiers in Bioscience 3; 2018

Computational tools can be used in the process of bringing a novel drug to the market to reduce cost and time using rational approaches. One of the main aim of using these tools is to choose between a number of small organic compounds, what ligands that shall bind to the biological target. Prediction of receptor-ligand affinities is of outmost importance in this stage of computer-aided drug design and still a challenge in computational biophysics. Moreover since chemical processes occur in an aqueous environment, the theoretical and computational modeling of them has to account for solvent effects to match experimental conditions. The use of quantum mechanical (QM) methods is a natural way to achieve a detailed description of the noncovalent interactions between the target receptor and the ligand, needed for the determination of receptor-ligand affinity. However, the accuracy of QM calculations in extended systems using continuum solvent models depends on the  availability of optimized atomic radii and surface tension coefficients. On this research, we aim to derive optimized sets of the COntinuum Solvent MOdel (COSMO) [1] radii and surface tension coefficients for different semi empirical Hamiltonians to reproduce hydration free energies of a set of small molecules for which experimental data are available. Finally, we use a semiempirical quantum mechanical based approach for calculating binding affinity in large biomolecular systems using those optimized parameters. This approach presented a good correlation with experimental binding free energy measured in the CDK2 protein structure in complex with 15 different inhibitors.