A hybrid modeling strategy using Nuclear Overhauser Effect data with contact information

Abstract

A hybrid modeling strategy using readily assigned Nuclear Overhauser Effect (NOE) data and main chain N-O contact information from homology template proteins is suggested. They are designed to be complementary. The strategy yields structures with reasonable root-mean square distance (RMSD) and NOE violation. When compared to homology-modeled structures, the hybrid-modeled structures are better in terms of RMSD, RMS NOE violation, and protein-like scores, although the number of used restraints is smaller than the homology modeling. The structure itself can be used as a theoretically modeled structure and also as a starting point for further NOE data assignment in structure determination.