Predicting virus Fitness: Towards a structure-based computational model.

Area of Science:

• Virology and Molecular Biology
• Computational Biology and Bioinformatics
• Immunology and Vaccine Development

Background:

• Emerging viral mutations pose significant challenges for public health surveillance and understanding pathogen evolution.
• The SARS-CoV-2 spike protein (S-protein) is crucial for host cell infection via ACE2 receptor binding and is also a target for neutralizing antibodies.

Purpose of the Study:

• To develop a computational virus fitness model predicting the impact of mutations on SARS-CoV-2.
• To identify mutations that enhance viral entry by altering the balance between ACE2 receptor binding and antibody neutralization.

Main Methods:

• Employed structure-based computation to assess the effects of approximately 380,000 possible S-protein mutations on binding to ACE2 and antibody complexes.
• Introduced ACE2-antibody selectivity change as a key metric for fitness, enabling error cancellation.
• Validated the model against experimental binding and antibody escape data.

Main Results:

• Developed a model that categorizes viral mutations based on their potential to increase selective binding to ACE2 over antibody capture.
• Demonstrated that mutations increasing ACE2 binding relative to antibody binding are predicted to become fixated.
• The model successfully predicted viral variant behavior based on binding affinities.

Conclusions:

• The proposed virus fitness model provides a framework for predicting the evolutionary trajectory of viral mutations.
• This approach can aid in understanding the molecular mechanisms driving viral evolution in hosts.
• Findings may inform the development of variant-specific vaccines and enhance viral surveillance strategies.