EvoEF2

EvoEF2: Energy Function for De Novo Protein Sequence Design

EvoEF2 implements an energy function framework with nine optimized energy terms for computational de novo protein sequence design, using sequence recapitulation rather than thermodynamic mutation data to parameterize the model.

Key Features:

• Energy Function Framework: Incorporates nine distinct energy terms optimized for protein sequence design.
• Sequence Recapitulation Optimization: Calibrates parameters using native sequence recovery instead of thermodynamic mutation datasets.
• Native Sequence Recovery: Recovers 32.5% of residues in 148 monomer proteins (47.9% core, 22.3% surface) and 30.9% in 88 dimer proteins (42.4% core, 31.3% interface, 21.4% surface).
• Foldability Prediction: Evaluates designed sequences using I-TASSER; 87.8% of 148 monomer designs show RMSD < 2 Å relative to native structures.

Scientific Applications:

• De Novo Protein Design: Designs protein sequences for monomeric and dimeric structures with assessment of structural foldability.
• Protein Interface Analysis: Evaluates residue recovery in core, surface, and interface regions of protein complexes.

Methodology:

EvoEF2 optimizes nine energy terms through sequence recapitulation by maximizing native residue recovery across monomeric and dimeric proteins. Designed sequences are structurally validated using I-TASSER, and fold similarity is quantified by root-mean-square deviation (RMSD) relative to native structures.