Analysis details

We regularly download and curate whole-genome data from GISAID. These data are passed through a SARS-CoV-2 analysis pipeline, which has not been published as a standalone tool, but has been used in a number of publications, including [1,2,3]. All sequences are partitoned into three-month intervals (by sampling dates) and subjected to a series of selection analyses. This notebook presents a temporal evolutionary history of selected sites, as characterized by the [FEL method]. During a given time interval, a site can be

Positively selected: natural selection may be promoting diversification (changes to different amino-acids) at this site.
Negatively selected: natural selection may be suppressing amino-acid variation at this site; the site is therefore evolutionary conserved at the protein level.
Evolving neutrally: all other cases; this includes cases where the FEL method lacks statistical power.

When a site is identified as evolving non-neutrally, a p-value from the likelihood ratio test is used to assess significance. These are reported in the plot. Note that p-values are not directly comparable across timepoints, because of different sample sizes and other factors.