XEFoldMine

XEFoldMine predicts and interprets early folding residues (EFRs) from protein sequences to elucidate sequence determinants and their correlation with secondary structure during initial protein folding.

Key Features:

• Interpretable Predictions: Employs a grey box methodology to predict EFRs and provide interpretable, sequence-level rules governing early folding events.
• Dataset Analysis: Analyzes three datasets—natural proteins, scrambled sequences derived from those natural proteins, and de novo designed proteins—to identify sequence determinants that influence early folding.
• Secondary Structure Correlation: Correlates identified EFRs with the secondary structures they adopt upon folding to link initial folding events to final structural configuration.

Scientific Applications:

• Protein folding mechanism studies: Identifies residue-level initiators of folding to support mechanistic investigations of protein folding pathways.
• Predictive model development: Provides interpretable EFR rules to inform predictive models of protein behavior and stability.
• Design and disease research: Informs synthetic protein design and studies of folding-related disease mechanisms by revealing sequence determinants of early folding.

Methodology:

Uses a grey box approach to predict EFRs from protein sequences, extract residue-level sequence determinants, and correlate identified EFRs with the secondary structures they adopt upon folding.