Intermolecular Interactions in Biophysical and Material Science Applications
George
Kaminski
Assistant Professor of Chemistry,
My research interests include a wide range of theoretical/computational chemistry problems. The main focus is on building methodology and techniques useable in a variety of more applied projects. Computer power available to the scientific community has grown dramatically over the last decade because of the unprecedented advances in technology. Yet the methods of utilizing these new abilities are not always up to date. For example, there are still no robust methods and software available to compute intermolecular (primarily hydrogen-bond) interactions energies with an accuracy of ca. 0.5 kcal/mol. At the same time (just as one example) it has been demonstrated that such differences in a protein-ligand binding energies can make all the difference between a successful drug and a totally ineffective one.
A special focus of my research is on methods for accurate calculations of many-body interactions, primarily electrostatic polarization. The left picture below demonstrates the tremendous importance that the dielectric properties of a medium have on macromolecular structure and energetics:
|
Changes of DNA structure
after a protein, simulated as a region of low dielectric constant, is
introduced. From JACS, 118, 3787, 1996. |
Controlled growing of bio-organic crystals under a
Langmuir film. From Chem.
Rev., 101, 1659, 2001. |
Success in building the new methodologies mentioned above is expected to allow to estimate intermolecular interactions within the 0.5 kcal/mol accuracy and thus to contribute greatly into the fast growing area of computer-aided drug design.
Another area of applied computational research, in which accurate representation of the electrostatic interactions, including polarization, is very important, is self-assembly of crystal and other structures under two-dimensional ordered systems (Langmuir films) floating on liquid-air surfaces (right picture above). It is possible, by varying the film structure, to control the emerging crystal (or a macromolecule) and thus to create compounds with programmed properties. Accurate computational predictions of results of such processes will greatly decrease time and other resources needed in these projects. But it is imperative that the techniques for computing inter- and intramolecular interactions are accurate enough and include an adequate way of explicitly treating many-body effects (polarization).