COVID-19 Records

Private channel session 521

75 messages over 2h 25m, 2020-10-30 – 2020-10-30.

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.

  1. 2020-10-30 14:43 Edward C. Holmes open PDF p.240
    Here are the sequencing dates that I am pretty bloody certain are correct:
  2. 2020-10-30 14:43 Edward C. Holmes open PDF p.240
    •Dec 25: Patient samples sent to (sequencing) Company A •Dec 26: Fragments of SARS-CoV-2 discovered by Company A •Dec 27: Full genome obtained by Company A •Dec 27, 28: Company A report to CDC and hospital in Wuhan •Dec 30: Company B sequence data from another patient •Dec 31: Company B's report (59 reads) "leaked" on social media
  3. 2020-10-30 14:43 Edward C. Holmes open PDF p.240
    I think company B is the BGI.
  4. 2020-10-30 14:44 Kristian G. Andersen open PDF p.240
    Yeah, that's kinda what I thought. There's no question these four runs were registered on 12/10 though - which iskinda peculiar.
  5. 2020-10-30 14:45 Kristian G. Andersen open PDF p.240
    I thought maybe BGI would have made a bunch of registrations (because I assume they're sequencing from patients frequently), but that wasn't really the case - as far as I can see, only these four ones.
  6. 2020-10-30 14:45 Kristian G. Andersen open PDF p.240
    E.g., this one: https://www.ebi.ac.uk/biosamples/samples/SAMEA6507890
  7. 2020-10-30 14:48 Kristian G. Andersen open PDF p.240
    Interesting list of collaborating institutions there too - I assume Weifeng is probably the person from Shandong.
  8. 2020-10-30 14:49 Edward C. Holmes open PDF p.240
    Yes, I can check with Weifeng. I don't understand it, but I can't believe those dates are correct.
  9. 2020-10-30 14:50 Kristian G. Andersen open PDF p.240
    The dates are correct, I think - I don't think those can be wrong, since they're all recorded by the ENA - not thesubmitter.
  10. 2020-10-30 14:50 Andrew Rambaut open PDF p.241
    Could ask Ewan Birney to look into it. Perhaps they have additional logs?
  11. 2020-10-30 14:51 Kristian G. Andersen open PDF p.241
    Yeah, probably. It _is_ kinda peculiar to me - these are also all young people.
  12. 2020-10-30 14:53 Edward C. Holmes open PDF p.241
    When do they add the sample collection date?
  13. 2020-10-30 14:53 Kristian G. Andersen open PDF p.241
    Once they submit the data
  14. 2020-10-30 14:54 Kristian G. Andersen open PDF p.241
    But the "creation date" is stamped in automatically - I don't think they'd have filled in any of the other information until they submitted the data - but that's a good question for Ewan
  15. 2020-10-30 14:55 Kristian G. Andersen open PDF p.241
    All the assemblies are on the CNGB too - but no automatic dates stamped. https://db.cngb.org/search/project/CNP0000881/ (submitter being "Weijun Chen" - but obviously a million people with that name)
  16. 2020-10-30 14:55 Edward C. Holmes open PDF p.241
    So, to be clear, what happened on 12/10? The BGI sent a file to EBI?
  17. 2020-10-30 14:56 Kristian G. Andersen open PDF p.241
    No, on 12/10 BGI created four data entries with EBI. Data was then later filled in. Typically that's not really how it's done.
  18. 2020-10-30 14:57 Kristian G. Andersen open PDF p.241
    So creation is 12/10, but upload is at a later date (I don't think we can know when - collection date is 12/26 and the data was released to the public 1/30)
  19. 2020-10-30 14:58 Kristian G. Andersen open PDF p.241
    Creating a BioSample for a pandemic virus two weeks before it's known that you have a patient with said virus is alittle strange - if the timeline is correct.
  20. 2020-10-30 15:00 Edward C. Holmes open PDF p.241
    The collection date sounds totally right. I don't understand why there is a file creation date 2 weeks earlier. Makes nosense at all. And why would BGI know about this? They are a commercial company that have nothing to do with Wuhan. WIV do their sequencing in-house I think.
  21. 2020-10-30 15:03 Kristian G. Andersen open PDF p.241
    That's the bit that's strange - because typically you don't create a BioSample before you have the data, unless you know the data is immediately forthcoming and you want to make sure there's no delay in submitting the data.
  22. 2020-10-30 15:04 Kristian G. Andersen open PDF p.242
    So if those BioSamples were created on, say 12/25, then that's totally what you'd expect - but we're talking two weeks before there was any (official) realization of a novel coronavirus.
  23. 2020-10-30 15:05 Edward C. Holmes open PDF p.242
    Yes, I don't get it at all. I'm contacting people now. But 2 weeks also sounds WAY too far in advance!
  24. 2020-10-30 15:06 Kristian G. Andersen open PDF p.242
    Yeah, definitely - very interesting.
  25. 2020-10-30 15:21 Edward C. Holmes open PDF p.242
    Company A is Vision Medicals. I don't believe for a nanosec that BGI knew there was a pandemic virus on Dec. 10th. That time-line makes zero sense with any known timeline, the epidemiological history of the outbreak, nor what was on Chinese social media. Nor would WIV have anything to do with BGI. The best time-line is here and confirmed my colleagues in China: https://apnews.com/article/3c061794970661042b18d5aeaaed9fae
  26. 2020-10-30 15:23 Edward C. Holmes open PDF p.242
    The idea that 2-3 weeks before all this it was known at a commercial sequencing company based in a different city seems untenable to me. Nor does it fit the time-line of the first patients seen by the doctors (sorry to include awikipedia link): https://en.wikipedia.org/wiki/Li_Wenliang
  27. 2020-10-30 15:24 Edward C. Holmes open PDF p.242
    However, I'm checking with colleagues there now.
  28. 2020-10-30 15:33 Kristian G. Andersen open PDF p.242
    One explanation is that BGI created these four BioSamples for a different purpose and then backfilled with the SARS2 data when they generated it. But I don't think there's any ambiguity that these were created on 12/10 - onlywe don't know for what purpose (so could have been for a novel coronavirus, could have been for something else -the metadata describing the BioSample very likely weren't entered until the data were deposited).
  29. 2020-10-30 15:35 Kristian G. Andersen open PDF p.242
    I would love to see the full submission history for these entries - Andrew, dig a little :wink:
  30. 2020-10-30 15:51 Kristian G. Andersen open PDF p.242
    Eddie, it's your paper... :wink: https://www.thelancet.com/article/S0140-6736(20)30251-8/fulltext I assume this is the same Weijun Chen who's the submitter of the data and is from BGI.
  31. 2020-10-30 15:52 Kristian G. Andersen open PDF p.242
    He'd know what's up with that early creation date.
  32. 2020-10-30 15:54 Kristian G. Andersen open PDF p.242
    If you had some suspected cases of novel coronavirus, Weijun would be the kind of guy you'd contact - he's been developing diagnostics for SARS too, so would make sense. I think there's probably an earlier timeline here that they may have been obscuring a little - which, honestly, wouldn't surprise me that much.
  33. 2020-10-30 16:22 Edward C. Holmes open PDF p.243
    Yes, I looked at the Lancet paper straight away. I've already contacted Weifeng. If that fails I'll contact Weijun Chen.
  34. 2020-10-30 16:25 Edward C. Holmes open PDF p.243
    The point is who is the 'you' in your question? Who did they work for? A doctor in Wuhan? What people would do isrun the standard panel of tests and if they failed perhaps contact a sequencing company like BGI. But if that were the case why did they not submit samples until 2 weeks later? My guess is that they were created for something else and then backfilled.
  35. 2020-10-30 16:25 Kristian G. Andersen open PDF p.243
    Yeah, it'll be interesting to see what they say - also, this is from the paper about the metadata [shared file(s): Screen Shot 2020-10-30 at 4.08.13 PM.png]
  36. 2020-10-30 16:26 Kristian G. Andersen open PDF p.243
    "My guess is that they were created for something else and then backfilled" Yeah, that's my suspicion too - although that _is_ a little unsual.
  37. 2020-10-30 16:28 Edward C. Holmes open PDF p.243
    But I think the other option is even more unusual...but I'm as keen to find out as you! At the very least we need acompelling answer for the trolls who are about to descend upon us...
  38. 2020-10-30 16:30 Kristian G. Andersen open PDF p.243
    Yeah, once the tinfoil hats get hold of this. :boom:
  39. 2020-10-30 16:31 Edward C. Holmes open PDF p.243
    But it also doesn't fit the WIV narrative as they don't do anything with the BGI.
  40. 2020-10-30 16:33 Edward C. Holmes open PDF p.243
    And the BGI are not even in Wuhan.
  41. 2020-10-30 16:33 Kristian G. Andersen open PDF p.243
    Yeah, I don't really see how any of this fits with WIV (other than it being close to PLA). It fits better with the Huanan market (since the whole "Line 2" narrative isn't all that stupid)
  42. 2020-10-30 16:34 Kristian G. Andersen open PDF p.243
    One thing "Steve" considers a potential hypothesis here is a "vaccine challenge trial" - which isn't totally idiotic. But we'd have known about some SARS2 related virus if that were the case.
  43. 2020-10-30 16:35 Edward C. Holmes open PDF p.243
    Raccoon dogs in the market would be so simple....yes, the vaccine trial is mad.
  44. 2020-10-30 16:38 Edward C. Holmes open PDF p.244
    We know there must have cases around at that time. All those posterior diversity stuff is complete BS though. That Wade guy from the NYT kept rattling on about that.
  45. 2020-10-30 16:40 Kristian G. Andersen open PDF p.244
    Yeah, we know the early cases occurred back in the beginning of December, so it's certainly not completely mad tothink that somebody was investigating using sequencing around that time.
  46. 2020-10-30 16:42 Edward C. Holmes open PDF p.244
    But very likely in that there is just no footprint of this anywhere else...anyway, I've put the feelers out and I'll update you as soon as I hear anything.
  47. 2020-10-30 16:45 Edward C. Holmes open PDF p.244
    Also having a chat with Tommy later so will be armed with more pangolies.
  48. 2020-10-30 16:45 Kristian G. Andersen open PDF p.244
    Don't forget to wear your new shirt!
  49. 2020-10-30 16:46 Kristian G. Andersen open PDF p.244
    ... and just so we have the paper saved here on Slack... [shared file(s): COVID Pandemic Began at PLA Hospital inWuhan.pdf]
  50. 2020-10-30 16:46 Edward C. Holmes open PDF p.244
    I've had quite a few preorders for that shirt!!
  51. 2020-10-30 17:05 Edward C. Holmes open PDF p.244
    The BGI people first made their data entry in China on Dec. 26th:
  52. 2020-10-30 17:06 Edward C. Holmes open PDF p.244
    shared file(s): Screen Shot 2020-01-30 at 8.50.31 PM.png
  53. 2020-10-30 17:06 Edward C. Holmes open PDF p.244
    shared file(s): Screen Shot 2020-10-31 at 11.03.59 am.png
  54. 2020-10-30 17:07 Edward C. Holmes open PDF p.244
    This time-line makes sense but it does show that George Gao must have known about it before Dec. 30!! I never believed that one for a second.
  55. 2020-10-30 17:09 Kristian G. Andersen open PDF p.244
    Yeah, this totally makes sense given the timeline. And yeah - George must have known. The ENA entries must have been created for some other purpose and backfilled. But it _is_ weird - because they only created those four on ENA and then nothing for a few weeks. SLACK_000459
  56. 2020-10-30 17:09 Edward C. Holmes open PDF p.245
    We'll get to the bottom of it. That also provided me with Weijun Chen's email address so I'm just contacting him now.
  57. 2020-10-30 17:09 Kristian G. Andersen open PDF p.245
    Okay, cool - I'm sure they're going to have a totally mundane explanation - like why RaTG13 was named that way cwink:
  58. 2020-10-30 17:09 Edward C. Holmes open PDF p.245
    On it.
  59. 2020-10-30 17:09 Kristian G. Andersen open PDF p.245
    (and one pretty critical thing against the bonkers theories in the paper - wouldn't WIV have been mentioned in theBioSample had the virus come from there....? :man-shrugging:)
  60. 2020-10-30 17:09 Edward C. Holmes open PDF p.245
    Totally!
  61. 2020-10-30 17:09 Edward C. Holmes open PDF p.245
    Tommy says there are rumours that people knew earlier. My boy told me this: "As for the entry date, I have not read the long document yet, but that submission date is possible. They all talked about the sampling might occur much earlier (from different occasions). Therefore, I wouldn't be surprised if the sample is collected around that time (Dec 10th, not too bad). And I wouldn't be surprised if they have "changed" the sampling date". Nothing officially. Atworse it means that the China CDC must have known earlier which will throw the shit badly on them, particularly George. May be not a bad thing. Does not impact origins in any way as I see it.
  62. 2020-10-30 17:09 Edward C. Holmes open PDF p.245
    Tommy has 6 SARS-CoV-2 sero-positive pangolins that are currently being metagenomically sequenced. The other news - which is obviously strictly confidential - is that someone in China (Tong?) has apparently obtained a MERS-like virus from a pangolin.
  63. 2020-10-30 17:09 Kristian G. Andersen open PDF p.245
    Very interesting, makes sense. Based on this, the most likely explanation is that they probably had the very earliest patients sequenced around Dec 10th and then somebody slammed the brakes. The samples we have that are dated12/26 are likely earlier then - and actually supported by the sequences themselves (one of these appear basal). The root is funky too - @Andrew Rambaut, I wonder if uncertainty in the dates around the very earliest samples could help explain that? Probably... (although, last I checked, the root is still funky if you remove all the dates).
  64. 2020-10-30 17:09 Kristian G. Andersen open PDF p.245
    In any case - this is going to course another shit storm for the Chinese they're going to have a very hard time cleaning up - let's hope it happens after the election....
  65. 2020-10-30 17:09 Kristian G. Andersen open PDF p.245
    The most basal sample being WH4 - this one: https:/Avww.ncbi.nlm.nih.gov/biosample/SAMEA6507893. The GPS coordinates on that points to a nursing home right next to the Line 2 main station (patient being a 39 old man). SLACK_000460 Very interesting - it appears that there might have been some early cases the world didn't know about... (or rather, cases they _did_ know about, but they occurred earlier)
  66. 2020-10-30 17:09 Edward C. Holmes open PDF p.246
    Yes, a shit storm if true. But that's just Let's see if I can anything 'official'.
  67. 2020-10-30 17:09 Edward C. Holmes open PDF p.246
    I really want to see that pango data!
  68. 2020-10-30 17:09 Kristian G. Andersen open PDF p.246
    Yup - and I'm sure you will before the rest of us! Pangolying idiot!
  69. 2020-10-30 17:09 Edward C. Holmes open PDF p.246
    Those t-shirts will be priceless.
  70. 2020-10-30 17:09 Kristian G. Andersen open PDF p.246
    I'll need to get a bunch for VGE next year!
  71. 2020-10-30 17:09 Edward C. Holmes open PDF p.246
    Dear Eddie, Thank you for your email. To be honest, I noticed the outbreak from the media in the morning of December 31, 2019. George knew the outbreak several hours earlier than me, midnight December 30, 2019 also from the media. In addition, as you know, what I do is to analyze the genome sequences from BGI and China CDC. I don't know what is going on with the information you sent. I guess it might be a typo. Anyway, I have contacted the collaborator from BGI to give a logical explannation. If necessary, I will tell this to George. Best wishes, Weifeng
  72. 2020-10-30 17:09 Andrew Rambaut open PDF p.246
    I think the GPS coordinates are truncated to two decimal places so I wouldn't make too much inference from them (assume +/- 1km).
  73. 2020-10-30 17:09 Andrew Rambaut open PDF p.246
    @kristian Andersen Limited edition - signed.
  74. 2020-10-30 17:09 Kristian G. Andersen open PDF p.246
    Interesting email from Weifeng. Importantly though - the dates can't be typos as they're automatically generated....
  75. 2020-10-30 17:09 Edward C. Holmes open PDF p.246
    Yes, I told him that and that it was important we get it resolved. George is a slippery bugger but Weifeng is totally straight and a very good guy. If he says he knew on the 31st then I believe him. I cant' say I believe George though. I'll update as I hear more.