Cluster expansion models for flexible-backbone protein energetics.

Area of Science:

• Computational biology
• Biophysics
• Protein engineering

Background:

• Protein structure prediction and design commonly use discrete models for side-chain conformations.
• Incorporating backbone flexibility enhances model accuracy and expands accessible sequence spaces for computational design, but increases complexity and computation time.

Purpose of the Study:

• To develop an efficient method for incorporating backbone flexibility into protein modeling and design.
• To demonstrate that backbone flexibility's influence on protein energetics can be implicitly modeled at the sequence level.

Main Methods:

• Application of the cluster expansion technique to convert structure-based energies into sequence-dependent functions.
• Testing the method on flexible-backbone models of four diverse protein systems: alpha-helical coiled-coil dimers and trimers, zinc fingers, and Bcl-xL/peptide complexes.

Main Results:

• Cluster expansion enables dramatic speed-ups in energy evaluation by creating sequence-based energy functions.
• Sequence-based models derived using cluster expansion showed low errors when compared to structure-based evaluations across all tested protein systems.
• The method provides a convenient functional form for analyzing and optimizing sequence-structure relationships.

Conclusions:

• Implicit treatment of backbone flexibility using cluster expansion is effective for protein modeling and design.
• This approach offers significant computational advantages, particularly for large-scale protein design tasks.
• The method shows considerable promise for advancing computational protein design by efficiently handling backbone flexibility.