A Hybrid Monte Carlo Scheme for Multibackbone Protein Design

Area of Science:

• Computational Biology
• Protein Engineering
• Biophysics

Background:

• Multistate protein design involves optimizing side chain mutations while exploring a limited set of backbone conformations.
• Achieving accurate Boltzmann sampling of both protein sequences and their conformations is crucial for reliable design.

Purpose of the Study:

• To develop and validate a hybrid Monte Carlo (MC) scheme for Boltzmann sampling in multistate protein design.
• To compute backbone conformational free energies and verify the sampling method's accuracy.

Main Methods:

• A hybrid MC scheme combining backbone conformational sampling with side chain rotamer adjustments.
• Derivation of the theoretical and practical forms for a Metropolis-like acceptance test.
• Computation of backbone conformational free energies for SH2 and SH3 proteins.

Main Results:

• The hybrid MC scheme successfully enables Boltzmann sampling of sequences and conformations.
• The derived acceptance test provides a method for evaluating conformational changes.
• Computed free energies for SH2 and SH3 proteins behaved as a state function, confirming accurate sampling.

Conclusions:

• The developed hybrid MC approach is effective for multistate protein design, ensuring accurate Boltzmann sampling.
• This method facilitates the computation of reliable conformational free energies.
• The findings validate the theoretical framework for advanced protein design strategies.