Correlated mutations contain information about protein-protein interaction

Area of Science:

• Molecular Biology
• Bioinformatics
• Structural Biology

Background:

• Protein-protein interactions are crucial for cellular function, requiring specific molecular complexes.
• The evolution of interacting proteins leads to constraints on their sequences, reflecting co-adaptation.
• Inter-residue contacts at protein-protein interfaces are essential for interaction specificity.

Purpose of the Study:

• To develop a method for predicting contacting residue pairs from protein sequences alone.
• To demonstrate that correlated sequence changes can identify protein-protein interfaces.
• To show the applicability of this sequence-based approach to inter-domain and inter-protein interactions.

Main Methods:

• Application of a method for detecting correlated changes in multiple sequence alignments.
• Analysis of interacting protein domains and two-domain proteins.
• Testing the method on haemoglobin and predicting contacts for heat-shock protein Hsc70.

Main Results:

• Correlated sequence changes in interacting proteins occur at positions near protein-protein interfaces.
• Sequence information from correlated changes is sufficient to identify correct inter-domain docking solutions.
• The method successfully predicted inter-domain contact regions for Hsc70.

Conclusions:

• Correlated sequence changes provide a powerful signal for identifying protein-protein interaction sites.
• This sequence-based approach offers a novel and widely applicable alternative to traditional structure-based docking methods.
• The findings open possibilities for predicting protein-protein interfaces without requiring structural information.