Viral infection elicits antibodies targeting residues on the surface of glycoproteins (spikes). Because of the high-density presentation of spikes on the viral surface, not all antigenic sites are targeted equally by antibodies. We used molecular dynamics simulations to estimate antibody pressure/antibodies targeting of surface residues of the SARS-CoV-2 spike.
Antibody targeting of the spike (S protein) of SARS-CoV-2
For circulating viruses to propagate, they have to evade recognition by antibodies. They do so by accumulating mutations in the spike protein. We computed the mutational pattern for the sarbecovirus subgenus spike (SARS family) spike
Mutability map: entropy of the sarbecovirus subgenus spike residues. Low entropy=conserve, high entropy=mutable
To compare the affinity map and the mutability map we applied spectral clustering. We aggregated closeby residues to define epitope clusters.
How good is the model at explaining the mutability of surface residues?
The correlation between the affinity map and the mutability map is 0.69 for the sarbecovirus subgenus spike. To create the affinity map we use only the surface geometry of the virus and the spike. Hence, the virus surface geometry gives rise to antibody pressure, as estimated by the model, that can explain 48% of the variability in the mutability map.
Does SARS-CoV-2 surface geometry shape its spike mutability through antibody pressure?
We compared the mutability of the SARS-CoV-2 spike to the affinity map above. The spike acquires different mutations that can affect its infectivity or help it escape from antibodies. Over time its evolutionary tendencies may change. We explore its mutability map at different time periods (data from gisaid). Choose time window to pick mutations acquired in this period
Comparing the mutability map at this time window to the antibody pressure model
High correlation means that the relative mutability of spike residues follows their relative accessibility to antibodies and the geometry of the viral surface, as estimated by the model. It could be indicative of escape from antibody mutations, amongst other factors (see manuscript).
