Thread by @shay_fleishon

I see a lot of Omi samples with Delta mutations recently. Maybe contamination maybe convergence.
I went to look at it.
I took all non UK Omi samples from 26.12 (~17.5K) and looked for ones with Delta mutations (21j or 21i) using a code by @DrorBendet. I found 890 samples.

Here we can see that most of the samples are with 1-3 delta mutations. As we go to higher #delta mutations the probability for this to be a lab contamination is going up.

Here are the # of samples with each mutation. What pops up immediately is that the most prevalent ones are the more converged with other variants.

Here I omitted the samples with more than 2 delta mutations.

This is the S:L452R samples in the covariants omicron tree

Here is the M:I82T in the covariants tree

So, maybe there is some convergence. In several mutations that could be.
But also, @netazuckerman reminded me about the times when Delta took over alpha and pseudo “recombinants” rose from lab contaminations and that could lead now to some noise in the data (for a week or two)

Also, it would be nice to see this analysis in regard to other variants, and also to a list of mutations that I made which converge. What do you think ? worth the efforts ?

All this analysis of folks like me is a direct consequence of @GISAID which open this data to the world.

Thanks @Teitleraz , @DorYadin and Boris Boltiansky who helped with the data analysis.