1/ Medium thread on #SARSCoV2's furin cleavage site and a strikingly similar region in some of the new BANAL genomes from Lao, and in RmYN02 from China.
Seems worth trying to clear up the confusion of @ydeigin on this issue (even it means broadcasting my pic, below).
Seems worth trying to clear up the confusion of @ydeigin on this issue (even it means broadcasting my pic, below).
https://twitter.com/ydeigin/status/1443816256222728194
2/ The furin cleavage site of SC2 is the RRAR in the NSPRRAR stretch of amino acids in the alignment I'm holding up there. It is what makes the virus 'pop' in humans.
The BANAL viruses have NSPAAR. A couple other ones, including RmYN02, have NSPAAR or NSPVAR.
The BANAL viruses have NSPAAR. A couple other ones, including RmYN02, have NSPAAR or NSPVAR.
3/ So, having barely scratched the surface of the genetic diversity of these viruses in the wild, we've found several that are literally a *single* amino acid away from having a furin cleavage site.
For example: (NSP) ->inserted R<- AAR.
For example: (NSP) ->inserted R<- AAR.
4/ Regardless of how these genes are aligned with one another on our computers, this is true.
It is also clear that this region of the Spike protein is *highly* prone to insertions and deletions.
It is also clear that this region of the Spike protein is *highly* prone to insertions and deletions.
5/ Bat populations are like a vast bank of slot machines, and nature is pulling those arms down over and over, day in and day out, coming up with endless variations.
It is likely that a slight tweak of something like a BANAL Spike gave rise to RRAR in a bat.
It is likely that a slight tweak of something like a BANAL Spike gave rise to RRAR in a bat.
6/ Though of no relevance in the bat, that motif allowed that virus to hit the jackpot when it got into humans, most likely through intermediate hosts like raccoon dogs sold in wet markets.
7/ That's the big picture. What about the minor question of how these stretches of the spike gene should be aligned?
Yuri's proposal is possibly the least parsimonious scenario I have ever seen to explain a stretch of nucleotides.
Yuri's proposal is possibly the least parsimonious scenario I have ever seen to explain a stretch of nucleotides.
8/ By my count, it would require 10 mutations in an 18-nucleotide stretch. This is more differences than expected if you compare two randomly generated stretches of 18 nucleotides! Plus (for the aficionados), 5 would be transversions.
9/ Seems a lot less likely than a couple 'indels' in a region with lots of recombination.
10/ Again, no matter how a BANAL virus genome is aligned against others, that Spike protein contains NSPAAR, and is just one amino acid insertion away from a minimal furin cleavage site.
11/10 As @arambaut has pointed out, it is possible that NSPRRAR and NSPAAR are not homologous. But with respect to the origin of a furin cleavage site, homology does not matter.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
