Thread by @franciscodeasis

@MonaRahalkar @Daoyu15 @babarlelephant @Drinkwater5Reed @idomyownexpts @AntGDuarte @Rossana38510044 @jhouse678 @daoyu But we would know a few things:
- ancestor of Ra7896, so, before May-15
- very close to Ra4991, but not Ra4991 (has 5 mutations in 370)
- not the miner's virus
It could be what you mentioned long time ago: other samples. But not initially hidden, but detected very late. Maybe

@MonaRahalkar @Daoyu15 @babarlelephant @Drinkwater5Reed @idomyownexpts @AntGDuarte @Rossana38510044 @jhouse678 @daoyu OTOH, the RBD should come from the mine too. We already know there was one there in April 2012. They eventually found a descendant

@MonaRahalkar @Daoyu15 @babarlelephant @Drinkwater5Reed @idomyownexpts @AntGDuarte @Rossana38510044 @jhouse678 @daoyu A good candidate for WIV15 could be 1 of the 43 samples of Ra from the 3rd trip in April 2013

@MonaRahalkar @Daoyu15 @babarlelephant @Drinkwater5Reed @idomyownexpts @AntGDuarte @Rossana38510044 @jhouse678 @daoyu IIRC, there was a mismatch of 1 sample between the Table 1 and 2 of Ge et al. (2016). And could be due to that... Need to review this

@MonaRahalkar @Daoyu15 @babarlelephant @Drinkwater5Reed @idomyownexpts @AntGDuarte @Rossana38510044 @jhouse678 @daoyu Data provided by WIV is chaotic and inconsistent (as usual). I guess the correct thing is to assume there is 1 Ra missing in April 2013. I always thought it was due to the cheating of Ra3755, but it could be they wanted to hide an isolate (similar to 4991)

https://twitter.com/franciscodeasis/status/1272173870708133890

@MonaRahalkar @Daoyu15 @babarlelephant @Drinkwater5Reed @idomyownexpts @AntGDuarte @Rossana38510044 @jhouse678 @daoyu We always have the same problem: we need to separate genuine errors (a lot), academic cheating and cover ups. Everything superimposed. Not an easy task

@MonaRahalkar @Daoyu15 @babarlelephant @Drinkwater5Reed @idomyownexpts @AntGDuarte @Rossana38510044 @jhouse678 @daoyu You could be right according to this new theory. I did not see the point of hiding something with (almost) same sequence, but the point was instead that they may be hiding an isolate, which was useless due to a bad RBD, but that had an easy solution... :-(

https://twitter.com/MonaRahalkar/status/1281622187934277633

@MonaRahalkar @Daoyu15 @babarlelephant @Drinkwater5Reed @idomyownexpts @AntGDuarte @Rossana38510044 @jhouse678 @daoyu Imagine my hypothesis is right: SARS-CoV-2 comes from an infectious clone, whose backbone is very close to Ra4991. What would you see in a simplot? Here a real example comparing with the exact backbone (based on Menachery et al. (2015) infectious clone)

@MonaRahalkar @Daoyu15 @babarlelephant @Drinkwater5Reed @idomyownexpts @AntGDuarte @Rossana38510044 @jhouse678 @daoyu After outbreak is known, they are forced to publish complete genome of Ra4991 (partial RdRp published in 2016). So, how to fabricate a Ra4991 distant enough and hiding recombination while credible?: they limit themselves to use only real aa seqs (due to protein folding problem?)

@MonaRahalkar @Daoyu15 @babarlelephant @Drinkwater5Reed @idomyownexpts @AntGDuarte @Rossana38510044 @jhouse678 @daoyu The RdRp, a well conserved part, can be solved using Ra7896. Although some Ra79XX are probably more distant, Ra7896 is not far enough to contrast with the 370 nt of Ra4991. The RBD could be a 3rd RBD between Ra4991 and Ra79XX

@MonaRahalkar @Daoyu15 @babarlelephant @Drinkwater5Reed @idomyownexpts @AntGDuarte @Rossana38510044 @jhouse678 @daoyu What else you can do to hide the recombination?
- swap other segments with similar viruses when possible
- smooth jumps adding noise with synonym mutations
And in the simplot?
- remove gridline to hide peaks near 100%
- increase window size of simplot
- smash & stretch the graph

@MonaRahalkar @Daoyu15 @babarlelephant @Drinkwater5Reed @idomyownexpts @AntGDuarte @Rossana38510044 @jhouse678 @daoyu This could be the proof of the smoothing.
From the paper of @quay_dr.
zenodo.org/record/4477081…

@MonaRahalkar @Daoyu15 @babarlelephant @Drinkwater5Reed @idomyownexpts @AntGDuarte @Rossana38510044 @jhouse678 @daoyu @quay_dr It would be great is someone can draw a kind of simplot of:
1. dN/dS
2. percentage of codon coincidence (when aa matches)
I am not sure if this kind of plots exists, but should not be difficult to draw

@MonaRahalkar @Daoyu15 @babarlelephant @Drinkwater5Reed @idomyownexpts @AntGDuarte @Rossana38510044 @jhouse678 @daoyu @quay_dr This article is also very useful. The recombination is more evident.
advances.sciencemag.org/content/6/27/e…