We aim to develop first-principles theory for protein stability. The successes of first generation theory indicate that the prediction of protein stability from the amino acid sequence is now within reach. This is a necessary and important step on the much longer road to predicting protein structures from amino acid sequences. A framework which can incorporate and unify for the first time free energy contributions due to helix-coil, hydrophobic, electrostatic, and elastic and excluded volume entropic forces is proposed. The following problems will be attacked: (i) Zimm-Bragg-type theory accounts for polypeptide behaviors driven by """"""""local"""""""" interactions among neighbors in the chain sequence. Polymer collapse and folding is driven by """"""""nonlocal"""""""" interactions. To understand the full range of protein behaviors, we will incorporate both types of interaction into a single framework. (ii) The mean-field theory will be broadened to treat aspects of intermediate states and folding kinetics, (iii) Electrostatics will be (a) refined to improve predictions of the radius of highly charged unfolded states, (b) refined to account for discrete charge effects, and (c) applied to predicting protein aggregation. (iv) Side chain degrees of freedom will be incorporated into the entropy of folding. (v) The temperature dependence of polar group interactions will be taken into account, and effects of disulfides and other constraints on protein stability will be treated.
Our aim i s to produce theory and computer software that will predict the largeconformational-change properties of proteins. It will take as input the amino acid sequence, and produce as output phase diagrams which predict radii, solvent exposures, stabilities, charge state, aggregation state, and state transitions as functions of solution conditions.
