(emsl2471)Development and Analysis of the Hybrid Quantum Mechanics-Molecular Mechanics Approach in NWChem - Why Catalytic Antibodies Cannot Achieve the Level of Catalysis of Enzymes?
EMSL Project ID
2471
Abstract
To deepen our understanding of natural and synthetical processes and to support current experimental progress, there is an increasing need for affordable theoretical calculation of chemical reactions in complex systems like solutions, biomolecules and materials. This task is addressed by so-called hybrid quantum mechanical-molecular mechanical (QM/MM) methods, which build a bridge between accurate quantum chemical and efficient force field computations. Their treatment in NWChem was not satisfactory up to the latest release. However, very recently a new QM/MM model based on pseudobonds was implemented into the program. Thereby, it is intended to enhance the fidelity of the QM/MM module and to make its application more popular. In the first part of the proposed project, I will perform tests of this implementation on small model systems, with the focus on enzyme catalysis. After validation of the code, I will apply the QM/MM capabilities of NWChem to a large scientific problem, in which the question is addressed why catalytic antibodies cannot catalyze reactions as efficient as natural enzymes. Catalytic antibodies are currently the most promising enzyme mimetics, as they can be produced tailor-made for a given reaction. Application areas range from organic chemistry to human medicine. One possible application is the detoxification of contaminants, which makes this an important area for the Department of Energy. But although considerable effort was put into the elucidation and optimization of their properties, they cannot yet catalyze reactions with the same rate enhancements as typical enzymes. I hope to gain insight into the reasons for this drawback from a picture at the molecular level. Therefore, I plan to apply QM/MM computations to a prototype for the esterolytic class of antibodies, 17E8. Its mechanism of action is still under discussion, with links to the well-known catalytic triad of the serine proteases. The calculations will follow the different suggested reaction paths to compare their probability. The overall project is intended to take advantage of the excellent computing facilities at PNNL and will need processing time on the IBM-SP machines. It will be realized in cooperation with T. P. Straatsma and E. R. Vorpagel at EMSL.
Project Details
Project type
Capability Research
Start Date
2002-03-04
End Date
2003-09-30
Status
Closed
Released Data Link
Team
Principal Investigator
Team Members