FilterDCA

FilterDCA predicts inter-domain and inter-protein residue-residue contacts by integrating Direct Coupling Analysis (DCA) coevolutionary scores with averaged structural contact patterns to inform three-dimensional protein structure and complex assembly.

Key Features:

• Supervised Prediction Approach: Employs a supervised framework that integrates coevolutionary sequence analysis with structural information to produce interpretable contact predictions.
• Integration of Contact Patterns: Incorporates averaged contact patterns influenced by secondary structure into the prediction process alongside coevolutionary scores from Direct Coupling Analysis (DCA).
• Improved Performance with Transparency: Combines averaged contact patterns with DCA-derived scores to enhance predictive accuracy over standard coevolutionary analyses while maintaining interpretability.

Scientific Applications:

• Inter-domain and inter-protein contact prediction: Predicts residue-residue contacts across domains and between proteins to inform structural modeling of interfaces.
• Protein complex assembly modeling: Supports inference of three-dimensional assembly arrangements for multi-protein complexes from sequence coevolution and contact pattern information.
• Drug design: Provides interpretable contact information that can guide identification of interaction hotspots for small-molecule or biologic targeting.
• Functional annotation and evolutionary analysis: Uses coevolutionary and contact-pattern signals to aid functional interpretation of residues and to study evolutionary constraints on interactions.

Methodology:

Supervised integration of DCA-derived coevolutionary scores with averaged structural contact maps that capture secondary-structure–dependent contact patterns.