search

dms-vep / SARS-CoV-2_XBB.1.5_spike_DMS

Other
5 stars 1 forks source link

Assign distance to ACE2 to binding correlations #79

Closed Bernadetadad closed 6 months ago

Bernadetadad commented 6 months ago

@jbloom this addresses issue #75 . I simply added a toggle to binding correlation plots that allows to show correlation for sites proximal or distant from ACE2 or non-RBD sites.

The key takeaway is that indeed, ACE2 proximal site binding correlates much better than distal or non-RBD site binding for BA.2 vs XBB.1.5 lenti libraries. This may suggest that mutation RBD modulation effects on the two backgrounds are not equal.

Furthermore, for both BA.2 and XBB.1.5 lenti libraries correlations to yeast libraries increase when only ACE2 proximal sites are compared but for ACE2 distal sites there's basically no correlation to yeast RBD measurements.

jbloom commented 6 months ago

I'm closing this and merging identical changes in #81.

The only difference is I ran with --rerun-triggers mtime to make the pull request in #81 only change the compare_binding output and not re-run a bunch of other notebooks to re-generate docs files (I can show you this if interested in future). This just makes it easier to keep track of which files changed.