I am currently working on an extension of our molecular mechanics force field. Specifically, I am introducing formal lone-pairs into the molecular description and assessing the effect this has on the energetic and dynamic behavior of the system. In order to calibrate properly the parameters associated with these new """"""""atomic"""""""" centers, high level ab initio calculations have been performed on numerous small molecule systems. These results form the basis for comparison with the equivalent molecular mechanics calculations. In addition, free energy perturbation calculations have been carried out which provide a stringent assessment of the new molecular description. At this time, specific conclusions have not yet been drawn; however, a number of general effects have become clear. In all cases studied, inclusion of off-center charge sites improves the charge description. Specifically, the atom-centered charge set derived more faithfully reproduces the electrostatic potential and molecular dipole moment calculated quantum mechanically. In some cases, this improvement was dramatic and this was used as a first indication of where improvement was most necessary. In all cases, the chemically intuitive lone pair position was preferred although a few anomalous situations are still being investigated. In most cases, interaction energies between molecules were enhanced and greater directionality imposed. That is, conformational changes away from optimized structures more accurately reflects the quantum mechanical results. And finally, the free energy results indicate that molecules with lone pairs are more hydrophilic than when the extra centers are not present. These results have been submitted for publication to the Journal of Computational Chemistry. Currently, my research efforts are concentrated on applying this methodology to problems of more significant complexity. Specifically, I am studying the conformational energetics of the alanine dipeptide and alanine pentapeptide, as well as nucleic acid base pair interactions. The resources provided by the UCSF CGL are used to prepare input molecular structures and examine resulting output in order to insure proper model development. Without visual examination of both input and output, this project would be significantly more difficult.
Showing the most recent 10 out of 508 publications
