Efficient rotamer elimination applied to protein side-chains and related spin glasses

Folded proteins and spin glasses share various properties, such as seemingly random interactions between residues (spins), and one might presume that some generic behaviors of spin glasses would also be exhibited in a general way by proteins. But a comparison here shows that the side-chain conformation systems of apo-myoglobin and lysozyme are qualitatively different from specific closely related spin glass systems. This difference is manifest in the number of rotamers that can be identified as definitely not contributing to the global energy minimum. This identification is effected by using a significantly enhanced version of the Dead End Elimination theorem (Desmet, J., M. De Maeyer, B. Hazes, and I. Lasters. 1992. The dead-end elimination theorem and its use in protein side-chain positioning. Nature. 356:539–542), which is much more effective and efficient in eliminating rotamers. In several cases (for proteins, although not for spin glasses) this improved Dead End Elimination theorem succeeded in identifying the absolute global minimum of rotamer conformations, with no statistical uncertainty. The difference between protein and spin glass is due to correlations between the interactions of one residue pair with another pair, and probably will play an important role in the thermodynamic behavior of the protein system