One of the tenants of the structural genomics initiative is that structures of a large numbers of homologous proteins can be predicted computationally from a small number of experimental structures. The sets of homologous proteins to be predicted can differ substantially (by as much as 70%) in sidechain composition, making the structure of the protein backbone of prime importance in any modeling exercise. Determining a backbone structure directly, without inclusion of sidechains could also accelerate the structure determination process. However, this approach is not necessarily compatible with the usual NOE-based determination of solution structures by NMR. As a result, we have investigated determination of structures based primarily on angular constraints from residual dipolar couplings (RDCs). This approach will be described, and applications to small proteins selected as structural genomics targets will be used to illustrate the process.
Back to NMR: Application to Structural Genomics I
Back to The 56th Southeast Regional Meeting 2004 (November 10-13, 2004)