A Systematic Survey of Close Contacts in the Protein Data Bank

Helena Qi, Massachusetts Institute of Technology

Enzymes catalyze a variety of reactions, and studying them leads to deeper understanding of how proteins have solved the problem of controlling and enhancing reaction rates. Proteins are typically too large to effectively study using quantum mechanics-based methods, and often are studied with molecular mechanics methods instead. In molecular mechanics, atomic interactions are modeled using classical mechanics and parameterized force fields. As a result, simulations of proteins are unlikely to explore configurations where non-bonded atoms are in close contact, as they are high in energy. Effects like charge transfer and polarization, captured by quantum mechanics but not molecular mechanics, could increase the stability of the close contacts. However, a systematic survey of the Protein Data Bank, a repository of experimentally determined protein structures, reveals cases where they contain close contacts. With a 2.0 A-resolution cut off, the structures are of such high quality that the close contacts are likely a result of actual close contacts in the protein and not just experimental error. Applying quantum mechanical methods to study the interaction energies of these close contacts can lead to insights on why they form.

Abstract Author(s): H. Qi, H. Kulik