A deterministic global optimization approach for the protein folding problem

Abstract

A deterministic global optimization algorithm is proposed for locating the global minimum potentialenergy conformationsof oligopeptide chains. The ECEPP/3 detailed potential energy model is utilized to model the energetics of the atomic interactions. The minimization of the total potential energy is formulated on the set of peptide dihedral angles. Based on previous work on the microcluster and molecular structure determination , a procedure for deriving convex lower bounding functions for the total potential energy function is utilized which involves a number of important properties. The global optimization algorithm BB which has been shown to be {convergent to the global minimum potential energy conformation through the solution of a series of nonlinear convex optimizationproblems is utilized. The ECEPP/3 potential model is interfaced with BB in the program GLOFOLD, and provisions have been made to accommodate user speciied partitioning of the dihedral angles into three sets. The rst one (i.e., global variables), consists of dihedral angles where branching occurs. The second set (i.e., local variables) includes the dihedral variables where branching is not necessary. The third set, (i.e., xed variables) includes the dihedral angles which are kept xed. The proposed deterministic global optimization is applied on a number of oligopeptide folding problems.