Constrained evolutionary funnels shape viral immune escape.

Area of Science:

• Virology
• Evolutionary Biology
• Computational Biology

Background:

• Understanding viral protein adaptation under immune pressure is key to predicting antibody-resistant variants.
• Viral evolution involves navigating vast mutational landscapes while maintaining protein function.

Purpose of the Study:

• To develop a probabilistic framework for predicting viral escape trajectories.
• To identify and characterize viable evolutionary pathways for immune evasion.
• To apply the framework to SARS-CoV-2 to understand variant evolution and antibody effectiveness.

Main Methods:

• Developed a probabilistic framework using a generative model trained on homologous sequences and deep mutational scanning data.
• Modeled constraints of protein viability and antibody escape.
• Derived a mean-field approximation of evolutionary path ensembles to quantify fitness and entropy.

Main Results:

• Immune evasion is channeled into a limited number of viable "escape funnels" within the mutational space.
• The framework successfully predicted mutation sites in SARS-CoV-2 variants of concern.
• Revealed convergent evolution patterns in viral adaptation.
• Explained differential effectiveness of antibody cocktails.

Conclusions:

• Viral adaptation to immune pressure follows predictable, constrained pathways.
• Antibody cocktails with decorrelated escape profiles are more effective at slowing viral adaptation.
• The framework provides a tool for anticipating future viral variants and designing effective countermeasures.