Membrane proteins comprise about 30% of the proteins coded by the genome and about 60% of the therapeutic targets for clinical drugs, however less than 1% of the structures of the Protein Data Bank1. There is a relative lack of information about membrane protein dynamics compared to globular soluble proteins2. The experimental situation has improved a lot during the last few years, but the basic principles of membrane protein folding are still unknown3. The relative role of kinetical and thermodynamical factors as well as chaperone activity are still hotly debated. In the late 1980's Wolynes and Bryngelson developed a quantitative energy landscape theory to address the statistical physics of the globular protein folding problem4. The theory postulates that proteins have evolved to make more consistently stabilizing interactions in their native state than in other compact configurations5. These ideas not only have been met with success in the structure prediction field, but also have provided a new framework through which to understand folding and function of globular proteins more generally. I plan to exploit my experience in the Wolynes group to develop new tools for studying membrane protein folding computationally, for understanding in-vitro and in-vi vo membrane protein folding. I plan to start by addressing basic questions using structure based models (go models) in which only residue pairs that are in contact in a "native structure" interact6. I also plan to use a related energy landscape idea to try to treat non-native interactions realistically and en passant to try to predict membrane protein structures using a physical simulation model7. The broader impact of this project is to extend the energy landscape theory to the folding of proteins in biological membranes and to answer how do topological and energetic frustration enter? Is this control completely kinetic? This project will make computational tools available to the scientific community that will aid in predicting the kinetics of membrane protein assembly and predicting membrane protein topology.