Prediction of folding rates of small proteins: empirical relations based on length, secondary structure content, residue type, and stability.

Delineating the determinants of folding of small proteins is essential for decoding the folding code. By considering the literature data for the folding of 45 two-state proteins that differ widely in sequence, structure, and function, this work suggests two empirical relations for the prediction of the folding rate. One relation is based on the content of secondary structure elements, and the other uses the number of residues of each of the following types: hydrophobic, positively charged, and negatively charged. Both relations incorporate the chain length as an indirect descriptor. The correlation between experimental values for folding rates and free energy is poor, providing little support to the "stability gap" hypothesis based on the folding of simple lattice and off-lattice polymers. The average rate-determining barrier height for these two-state proteins is much larger than the small barriers for the transition-state ensemble envisaged by theoretical models.