UEP

UEP predicts the impact of single protein residue mutations on protein-protein interactions by classifying mutations as beneficial or detrimental.

Key Features:

• Three-body contact potential: Evaluates mutation effects using a three-body contact potential derived from interactome datasets.
• Classifier trained on interactome data: Uses a classifier trained on interactome data to assign beneficial or detrimental labels to mutations.
• Consensus selection: Combines UEP predictions with pyDock and EvoEF1/FoldX via a consensus procedure to improve classification accuracy.
• Algorithmic simplicity: Implements a straightforward model based on three-body contacts while maintaining competitive performance against established methods.
• Computational speed: Prioritizes computational efficiency to enable rapid prediction of mutation impacts on protein interactions.

Scientific Applications:

• Protein-protein interaction analysis: Predicts how single-residue mutations alter binding affinity within protein complexes.
• Drug design and disease modeling: Supports evaluation of mutation effects relevant to drug design, disease modeling, and therapeutic intervention studies.
• Protein engineering: Identifies mutations that enhance or disrupt specific interactions for protein engineering applications.

Methodology:

UEP applies a classifier trained on interactome data and a three-body contact potential to analyze changes in binding affinity caused by single-residue mutations and optionally integrates predictions from pyDock and EvoEF1/FoldX through a consensus selection procedure.