Private channel session 1607

36 messages over 1h 39m, 2021-09-21 – 2021-09-21.

A “conversation” here is an activity session — a run of messages with under 60 minutes of silence inside it. The channel had no native conversation boundaries.

2021-09-21 08:01 Kristian G. Andersen open PDF p.999

IMO we really _do_ need to take a close look at what's described in this grant and look back at our (now expanded)alignments to see if anything pops up. E.g., Bob mentioned that there are differences in Laos bat vs SC2 when itcomes to N-linked glycans, which isn't particularly reassuring - that's the kind of work described on page 11.
2021-09-21 08:05 Andrew Rambaut open PDF p.999

I agree - need to stay on top of this.
2021-09-21 08:05 Kristian G. Andersen open PDF p.999

Yes - this can't (shouldn't) be dismissed out of hand.
2021-09-21 08:06 Andrew Rambaut open PDF p.999

If there is anything suspicious - better that we say it and then address it. If there isn't then we can let it blow over.
2021-09-21 08:06 Kristian G. Andersen open PDF p.999

Exactly. I suspect it'll be the latter.
2021-09-21 08:08 Andrew Rambaut open PDF p.999

Me too. But it really isn't beyond plausibility that someone went ahead and tried some of this stuff. We just need tore-convince ourselves it is coincidence.
2021-09-21 08:10 Kristian G. Andersen open PDF p.999

Correct - that's the problem. There's a big difference between "the lab was working on bat CoVs" to "the lab wasmodifying bat CoVs using features we see being unique to SC2".
2021-09-21 08:10 Andrew Rambaut open PDF p.999

Key question is - would anyone chose to put the FCS we see in SC2 in and put it out of frame.
2021-09-21 08:11 Kristian G. Andersen open PDF p.999

Would be a very strange way to do it, but not impossible - depending on their cloning strategy.
2021-09-21 08:12 Andrew Rambaut open PDF p.999

If you wanted to insert an FCS you would pick the AAs you wanted, pick the codons you wanted and then insert it.
2021-09-21 08:12 Andrew Rambaut open PDF p.999

The alternative is that it was inserted in frame and then mutated to make it look out of frame
2021-09-21 08:14 Kristian G. Andersen open PDF p.999

Again, it depends on cloning strategy - if the use restriction enzymes and there would be an easy way to insert anFCS with PCR utilizing REs already present in the sequence, then it could be a way to do it. Not the optimal way, for sure, but sometimes ease trumps perfection.
2021-09-21 08:16 Andrew Rambaut open PDF p.999

Would you pick CGG for example? Would you consider what codon to use for an R or just pick one arbitrarily.
2021-09-21 08:21 Kristian G. Andersen open PDF p.1000

CGG would be an unusual choice IMO - not a 'optimized' R and sorta middle-of-the-pack as it comes to usage in human genomes. Very rare, of course, in viruses because it's a CpG.
2021-09-21 08:24 Andrew Rambaut open PDF p.1000

And you have 6 codons to chose from.
2021-09-21 08:25 Kristian G. Andersen open PDF p.1000

Correct. Key here too is that if you were to 'optimize' the FCS sequence (as is commonly done for these types ofexperiments) you would _not_ end up with the FCS sequence we see, including the CGG codons. But again, itdoesn't really tell us much.
2021-09-21 08:26 Kristian G. Andersen open PDF p.1000

What I think we have to look _very_ closely at are all the specific sites they talk about in the grant and then compare across our alignments - is SC2 unique in those positions compared to, say, the BANAL viruses? Does it flip back toSARS1 at some of these positions, etc.
2021-09-21 08:26 Andrew Rambaut open PDF p.1000

Yup.
2021-09-21 08:27 Kristian G. Andersen open PDF p.1000

I'll get some of this started later today - it's important to get the insights irrespective of what it's going to end upshowing.
2021-09-21 08:28 Andrew Rambaut open PDF p.1000

Would there be any reason to pick the SARS2 backbone for doing these? Assuming they had the genome sequence way back at the start?
2021-09-21 08:30 Kristian G. Andersen open PDF p.1000

Well, in the grant they specifically talk about *new* viruses - so a SC2 precursor could have been one of those "low risk" viruses they talk about and then they started fiddling with it.
2021-09-21 08:34 Andrew Rambaut open PDF p.1000

They would have selected it from candidates by genome sequencing it. I guess they would have seen an ACE-2 binding RBD and thought that is the one to try an FCS in?
2021-09-21 08:37 Andrew Rambaut open PDF p.1000

Or they just were selecting viruses by seeing if they could get any to culture and then worry about genome sequencing?
2021-09-21 08:48 Kristian G. Andersen open PDF p.1000

The former - which is the approach they describe in the grant.
2021-09-21 08:49 Kristian G. Andersen open PDF p.1001

Culture is very challenging so I don't think that approach makes sense.
2021-09-21 09:00 Kristian G. Andersen open PDF p.1001

Here's what I suggest we do: • Make (spike) alignments with SC1, SARSr-CoVs, RaTG13(+close friends), Pangos, BANAL, SHC014, rs4237, SC2 • Mark 'key' sites based on what's in the grant - @Robert Garry, can you please help highlight the N-linked glycans based on Wuhan Hu-1 coordinates• Check to see if there's anything odd between SC2, basal viruses, and SC1• Look at e.g., codon usage at those sites if anything stands out• Cross-check with how those sites have evolved during the pandemic
2021-09-21 09:01 Andrew Rambaut open PDF p.1001

And stay off email
2021-09-21 09:02 Andrew Rambaut open PDF p.1001

Or carefully curate some emails for the FOI records
2021-09-21 09:07 Robert F. Garry open PDF p.1001

OK I'm on my bit...
2021-09-21 09:07 Kristian G. Andersen open PDF p.1001

Exactly - let's dig around a little and then start an email chain. I'm in Zoom calls most of the day, but I'll dive in in the afternoon.
2021-09-21 09:09 Kristian G. Andersen open PDF p.1001

Actually, let me get started - I just cleared my calendar for the next hour.
2021-09-21 09:34 Robert F. Garry open PDF p.1001

Is this a good set for the SC1 related viruses - or others (some more civet ones)? [shared file(s): image.png]
2021-09-21 09:35 Kristian G. Andersen open PDF p.1001

I'm getting (and annotating) the genomes now
2021-09-21 09:37 Andrew Rambaut open PDF p.1001

Ooh.. Can't wait.
2021-09-21 09:39 Andrew Rambaut open PDF p.1001

Probably worth having the full genomes aligned even if we only focus on the close relatives in the spike
2021-09-21 09:41 Kristian G. Andersen open PDF p.1001

Yup - I'm getting all the full length genomes first and then we can cut alignments out from there. Getting them annotated too - future Kristian will thank present Kristian.