Global optimization of protein 3-dimensional structures in NMR by a genetic algorithm

Protein three-dimensional structure determination is one of the most important problems in molecular biology. Nuclear magnetic resonance (NMR) spectroscopy is one of the promising techniques capable of determining the three-dimensional structures of proteins at atomic resolution. In determining protein structures using NMR spectroscopy, nuclear Overhauser effect (NOE) signal assignment is the most laborious and time-consuming process. Attempts to automate NOE signal assignment have failed so far. In this paper, we propose a new automatic assignment method of NOE signals based on a real-coded genetic algorithm and examine its effectiveness by applying it to determining the structure of an /spl alpha/-helix, which is a well-known common substructure of proteins, and a protein called HMG2B.