Next: Introduction Up: Dynamic and Stochastic Protein Previous: Conclusions

Chapter 4

Probability Grid Monte Carlo PGMC.


We have devised a Monte-Carlo method that employs importance sampling of dihedral angles to model peptide and protein conformations. This new method, which we call Probability Grid Monte Carlo (PGMC), modifies amino acid residue backbone and/or sidechain dihedrals according to probability grids derived from the Brookhaven Protein Database. We have used this method to study peptide conformations and have successfully adapted it to a number of important problems in protein modeling, including the prediction of all-atom protein conformations from C coordinates alone, and the prediction of the conformations of protein loops. Here, the method is applied to a study of the low energy conformations of the peptide Met-enkephalin.
Sat Jun 18 14:06:11 PDT 1994