Would we be even calling this a "variant" if we were not in the middle of the pandemic? If all you have is one sequence of SARS-CoV-2 and one sequence of this, sampled from bats, you probably call them "strains" and give them different designations...

Many of these S mutations aren't exactly common and quite a few are extremely rare. Aside from S371L which is unsurprisingly new since it is a 2nt mutation, all of N856K, Q954H, N969K, L981F have been seen fewer than 100 times. Q493K and Y505H were seen in the New York wastewater samples but are uncommon in humans (<200 samples). Lots more are rare enough that one wouldn't suspect they were advantageous. It's extremely unlikely that so many inconsequential mutations would accumulate in the spike rather than being more evenly spread through the genome, so the logical conclusion is that most of them are not inconsequential - even those ones from 856 to 981.

Compare this with B.1.640 for example, where even the more odd-looking polymorphisms like N394S had been seen at least several hundred times before. (F490R was new but again that's a 2nt mutation.)

I'm not saying this is spillback from an animal reservoir after a year of adaptations in that species, but if that were to happen, this is the sort of thing it might look like.

Edit: scratch that, the insertion is clearly human genome-derived. Unless that bit is shared with other mammals...?

Also, has any sequencing been done on the outbreak among students in Pretoria which is partly responsible for SA's recent rise in cases? It has all the signs of being driven by a new variant (very localised sudden rise in cases).

Additional sequence from South Africa: EPI_ISL_6647961

Futher sequences from South Africa (flagged by @shay671)
EPI_ISL_6647956
EPI_ISL_6647957
EPI_ISL_6647958
EPI_ISL_6647959
EPI_ISL_6647960
EPI_ISL_6647962

Edit: accidentially included the origional 4 sequences in this list when I posted it earlier - now corrected

Also, the South Africa ones all seem to be from Gauteng, Which is the province in which cases are rising right now...

Looks interesting, should definitely get designated soon @chrisruis

Interesting that these make up 100% of November-collected Gauteng sequences yet have at least slightly different location data around Johannesburg and a range of sampling strategies (one vaccine breakthrough, others surveillance), on the face of it suggesting high prevalence there

One of the patient (EPI_ISL_6647962) is 0 year old baby girl, meaning it's a household transmission.

Three of the sequences from South Africa have additional location information. Two of them are from Ekurhuleni (EPI_ISL_6647959 and EPI_ISL_6647956), and the other from Johannesburg Metro (EPI_ISL_6647957). These two places are ~70KM away from each other.

OR Tambo airport is in the Ekurhuleni region.

The case detected in Hong Kong only stayed in South Africa for ~20 days. He tested positive while being asymptomatic at the 2nd PCR test 3 days after returning to Hong Kong from South Africa, which means it's likely that he got infected just before boarding the returning flight - and OR Tambo airport is in Ekurhuleni.

Edit: scratch that, the insertion is clearly human genome-derived. Unless that bit is shared with other mammals...?

For what it's worth, the relevant section of the TMEM245 gene is identical in rhesus macaques (in fact it matches for one additional base!) but has a few differences in mice. So could theoretically be spillover from primates, but that seems less likely than an extremely prolonged infection in a human.

NSP4 – T492I

Interestingly this is the same ORF1b:T2163I which we've just heard about a few hours earlier in the context of #337 where it was seen as a unique addition to AY.43 potentially giving it a super-infectivity edge.

thx all for comments and additional info 🙏.
N679K was doing quite well in the last P.1 sublineages,

Thanks @thomasppeacock We've added this as B.1.1.529 in v1.2.100

I have locked the conversation on this issue as this Github is not intended for general discussion or speculation.

The original post initially had an incorrect mutation at Spike 493 - this should be Q493R and is now updated in the original post