PROTEIN-FOLDING SIMULATIONS WITH GENETIC ALGORITHMS AND A DETAILED MOLECULAR DESCRIPTION

We have explored the application of genetic algorithms (GA) to the det ermination of protein structure from sequence, using a full atom repre sentation. A free energy function with point charge electrostatics and an area based solvation model is used. The method is found to be supe rior to previously investigated Monte Carlo algorithms. For selected f ragments, up to 14 residues long, the lowest free energy structures pr oduced by the GA are similar in conformation to the corresponding expe rimental structures in most cases. There are three main conclusions fr om these results. First, the genetic algorithm is an effective method for searching amongst the compact conformations of a polypeptide chain . Second, the free energy function is generally able to select nativel ike conformations. However, some deficiencies are identified, and furt her development is proposed. Third, the selection of native-like confo rmations for some protein fragments establishes that in these cases th e conformation observed in the full protein structure is largely conte xt independent. The implications for the nature of protein folding pat hways are discussed. (C) 1997 Academic Press Limited.