Protein Folding

With the current knowledge, we can predict the primary sequence of amino acids in a protein from the DNA sequence of a gene. However, the key to protein function lies in the secondary and tertiary interactions between those amino acids that occur as a result of a process called protein folding. Protein folding takes a linear sequence of amino acids and creates a three-dimensional molecule with a defined structure. Within this defined structure lies the functional attributes of the protein. Therefore, if we can understand the process of attaining the folded state, we can predict the structure as well as the function of proteins from genomic sequence.

One direction we are exploring is the mechanics of proteins: ignoring the energy dynamics of the folding process, how can proteins reconfigure from one conformation to another? Here we model the protein as a mechanical linkage, either a chain representing the protein backbone or a tree representing all amino acids. Linkage reconfiguration has been studied extensively in geometry, but few positive results about reachability are known in 3D. The work exploits the special properties of proteins, e.g., that they are initially produced by ribosome, to understand why proteins can reconfigure without obstruction.

We are also looking at a variation on the standard protein folding problem: designing synthetic proteins that fold stably into a particular configuration. So far we have shown the existence of such stable foldings, i.e., proteins with unique minimum-energy states, in a mathematical model of protein folding called the H-P model. This model takes into account the hydrophobic and hydrophilic nature of protein segments and computationally determines the folded state