Evaluating the effect of SARS-CoV-2 spike mutations with a linear doubly robust learner

Area of Science:

• Virology
• Genomics
• Computational Biology

Background:

• Diverse SARS-CoV-2 variants emerge due to Spike protein mutations, prolonging the pandemic.
• Identifying key mutations driving viral fitness is essential for pandemic control.

Purpose of the Study:

• To develop a causal inference framework for identifying key SARS-CoV-2 Spike mutations.
• To evaluate the contribution of mutations to viral fitness and predict transmission capacity.

Main Methods:

• Applied causal inference methods to large-scale SARS-CoV-2 genomes.
• Estimated statistical contributions of mutations to viral fitness across lineages.
• Validated identified mutations computationally for functional effects (stability, binding, immune escape).

Main Results:

• Identified key fitness-enhancing mutations (e.g., D614G, T478K) and critical protein regions (RBD, NTD).
• Developed mutational effect scores to compute strain fitness and predict transmission capacity.
• Validated prediction model with BA.2.12.1, demonstrating accuracy.

Conclusions:

• This is the first study applying causal inference to SARS-CoV-2 mutational analysis on large genomes.
• Findings provide systematic insights into SARS-CoV-2 evolution and guide functional studies of key mutations.
• The framework enables reliable prediction of viral transmission capacity from genomic sequences.