Data-driven models for protein interaction and design.

Area of Science:

• Computational biology
• Structural biology
• Protein engineering

Background:

• The Critical Assessment of Predicted Interactions (CAPRI) is a community-wide experiment to assess computational methods for predicting the structure and properties of protein complexes.
• Advancements in computational methods are crucial for understanding protein interactions and designing novel protein functions.

Purpose of the Study:

• To introduce and evaluate four new challenge types within CAPRI, focusing on protein interface design and the impact of mutations.
• To assess the performance of computational methods in predicting protein-interface properties and binding affinities.
• To explore the role of interface-bound water molecules and heparin binding in protein interactions.

Main Methods:

• Development of novel prediction tasks for protein interface design, including distinguishing functional from non-functional interfaces.
• Prediction of the effects of single-point mutations on protein-protein binding affinity, using hemagglutinin as a model system.
• High-resolution structure prediction for interface-bound crystallographic waters.
• Modeling of heparin binding to a glycosylase protein.

Main Results:

• The study presents the methodologies and outcomes for the four new CAPRI challenges.
• Results offer insights into the capabilities and limitations of current computational tools for protein interface modeling and design.
• Performance metrics for predicting mutation effects and modeling water/ligand interactions are detailed.

Conclusions:

• The new CAPRI challenges provide valuable benchmarks for advancing computational protein design and interaction prediction.
• The outcomes highlight areas where further methodological development is needed in computational structural biology.
• This work contributes to the broader goal of understanding and engineering protein interactions for scientific and therapeutic applications.