Reduced representation approach to protein tertiary structure prediction: statistical potential and simulated annealing

A reduced representation model has been developed and used to predict the folded structures of proteins from their primary sequences and random starting conformations. The molecular structure of each protein is reduced to its backbone atoms (with ideal fixed bond lengths and valence angles) and each side chain approximated by a single virtual united atom. The co-ordinate variables are the backbone dihedral angles φ and ψ. A statistical potential function, which includes local and non-local interactions and is computed from known X-ray elucidated protein structures, is used in the structure minimization. Simulated annealing method of energy minimization is employed to search for folded conformations. Simulations using the reduced representation model reproduce many structural features of the studied proteins.