EvoEF2: accurate and fast energy function for computational protein design.

Area of Science:

• Computational Biology
• Protein Engineering
• Biophysics

Background:

• Current de novo protein design methods face limitations in accuracy and success rates.
• Parameter over-fitting in existing energy functions hinders discrimination between correct and incorrect protein designs.

Purpose of the Study:

• To develop an enhanced energy function, EvoEF2, for efficient de novo protein sequence design.
• To improve the accuracy and success rate of computational protein design.

Main Methods:

• Developed EvoEF2, an extended physical energy function building upon EvoEF.
• Optimized EvoEF2 parameters using sequence recapitulation.
• Evaluated EvoEF2 performance on monomer and dimer datasets for native sequence recovery.
• Assessed the foldability of designed sequences using I-TASSER.

Main Results:

• EvoEF2 achieved significant improvements in native sequence recapitulation for both monomers and dimers compared to EvoEF.
• Designed monomer sequences were predicted to fold into structures highly similar to native counterparts (87.8% with RMSD < 2 Å).
• Parameter optimization using sequence recapitulation proved more effective for computational protein design than using thermodynamic mutation data.

Conclusions:

• EvoEF2 offers a more accurate and efficient approach for de novo protein sequence design.
• The study highlights the critical role of parameter optimization in the effectiveness of physical energy functions for protein design.
• EvoEF2 demonstrates superior performance and applicability in designing novel protein sequences and interactions.