chapter  4
70 Pages


From Neural Networks to Protein Structure Prediction and Functional Dynamics of Complex Biomolecular Assemblies
WithGaregin A. Papoian, Peter G. Wolynes

This chapter reviews one particular class of useful coarse-grained protein models that has been developed over the last 25 years in the laboratories. The contemporary form of the resulting protein energy function is called the associative memory, water mediated, structure and energy model (AWSEM). The chapter elaborates on the details of the AWSEM potential and describes the parameter optimization strategy that follows from energy landscape theory. It discusses a number of applications of AWSEM to predicting the folding of globular proteins, and also to predicting mechanisms of protein association and formation of protein-Deoxyribo Nucleic Acid (DNA) complexes, such as nucleosomes. The chapter briefly reviews protein folding theory, clarifying how various fundamental ideas from the theories of spin glasses and polymer physics underlie the folding process. It discusses a series of coarse-grained protein force fields that have been used for predicting protein structure from amino-acid sequences either with or without knowledge about homologous proteins.