How to generate improved potentials for protein tertiary structure prediction: A lattice model study

Success in the protein structure prediction problem relies heavily on the c hoice of an appropriate potential function. One approach toward extracting these potentials from a database of known protein structures is to maximize the Z-score of the database proteins, which represents the ability of the potential to discriminate correct from random conformations. These optimiza tion methods model the entire distribution of alternative structures, reduc ing their ability to concentrate on the lowest energy structures most compe titive with the native state and resulting in an unfortunate tendency to un derestimate the repulsive interactions. This leads to reduced accuracy and predictive ability. Using a lattice model, we demonstrate how we can weight the distribution to suppress the contributions of the high-energy conforma tions to the Z-score calculation. The result is a potential that is more ac curate and more likely to yield correct predictions than other Z-score opti mization methods as well as potentials of mean force. Proteins 2000;41:157- 163. (C) 2000 Wiley-Liss, Inc.