SnugDock: paratope structural optimization during antibody-antigen docking compensates for errors in antibody...

Area of Science:

• Structural Biology
• Computational Biology
• Immunology

Background:

• High-resolution structures of antibody-antigen complexes are crucial for understanding binding interfaces and guiding antibody engineering.
• Predicting these structures computationally is necessary when experimental structures are unavailable.
• Existing methods often struggle with inaccuracies in antibody homology models.

Purpose of the Study:

• To introduce SnugDock, a novel computational approach for predicting high-resolution antibody-antigen complex structures.
• To demonstrate SnugDock's utility, particularly when antibody crystal structures are missing.
• To evaluate SnugDock's performance against standard docking methods.

Main Methods:

• SnugDock simultaneously optimizes rigid-body positions, antibody chain orientation, and complementarity-determining region (CDR) loop conformations.
• The method was tested using RosettaAntibody homology models and standard rigid-body docking for comparison.
• Ensemble docking was combined with SnugDock to simulate conformer selection and induced fit.

Main Results:

• SnugDock, particularly with RosettaAntibody homology models, yielded more accurate predictions than standard rigid-body docking.
• The combined SnugDock and ensemble docking approach achieved four medium and seven acceptable predictions in a set of fifteen complexes.
• While diverse paratope conformations were sampled, docked backbones did not always improve upon initial homology models.

Conclusions:

• SnugDock offers a significant advancement in predicting antibody-antigen complex structures, especially from homology models.
• The flexibility in optimizing CDR loops and other components enhances prediction accuracy.
• This work paves the way for a new class of flexible-interface docking algorithms to leverage homology models for high-resolution predictions.