--- type: Conversation title: Private channel session 1554 description: "slack conversation, 43 messages, 2021-07-02." tags: - conversation - slack medium: slack message_count: 43 started_at: "2021-07-02T08:29:54" ended_at: "2021-07-02T09:36:01" duration_seconds: 3967 sources: - { id: slack_pm, resource: /sources/slack-pm.md, title: Slack / Private Message Drop, author: "process:senate-hsgac-release" } generated: { by: randpaul-corpus-pipeline/1.0, at: "2026-08-01T12:42:27Z" } status: stable --- A slack conversation of 43 messages spanning 1h 6m. # Metadata - **Started:** 2021-07-02T08:29:54 - **Ended:** 2021-07-02T09:36:01 - **Duration:** 1h 6m - **Participants:** 2 - **Source:** [slack_pm](/sources/slack-pm.md) - **Pages:** 968-972 # Participants * [Andrew Rambaut](/people/andrew-rambaut.md) * [Edward C. Holmes](/people/edward-c-holmes.md) * [Kristian G. Andersen](/people/kristian-g-andersen.md) * [Robert F. Garry](/people/robert-f-garry.md) # Transcript **[2021-07-02T08:29:54] Andrew Rambaut:** Do we have an editable version of the figures? Do you want me to try to integrate them and style them a bit? **[2021-07-02T08:38:57] Andrew Rambaut:** I would like to do something with this but the resolution is too bad to extract the data even though it is just binary. Ican't see the exact districts in the centre. We need to get the raw data or better quality figures. **[2021-07-02T08:43:31] Andrew Rambaut:** Actually looking more carefully - I don't think it is much use to us **[2021-07-02T08:43:40] Kristian G. Andersen:** We unfortunately don't have either raw data or higher quality figures :disappointed:. I asked Marion, but she said there's no way to get the data since all of that is on the Chinese side. The excess deaths plot is just a fill within someneighborhood borders, so maybe we can find a vector of their maps? **[2021-07-02T08:43:49] Andrew Rambaut:** It is all in week 3 or later **[2021-07-02T08:44:15] Andrew Rambaut:** The December stuff I guess the numbers are too small to be detectible **[2021-07-02T08:44:28] Andrew Rambaut:** Shows it didn't start in August though **[2021-07-02T08:44:38] Kristian G. Andersen:** Yeah, everything is happening over three weeks in January, so those would be the ones that I think would be helpful **[2021-07-02T08:44:58] Kristian G. Andersen:** Having cases in December and then excess deaths in January would fit the bill **[2021-07-02T08:45:21] Andrew Rambaut:** But the spatial pattern by January would mean much I think **[2021-07-02T08:45:45] Kristian G. Andersen:** It's mostly that spatial pattern we refer to - because it starts North **[2021-07-02T08:46:20] Andrew Rambaut:** I guess we might just be able to work out the Districts: [shared file(s): image.png] **[2021-07-02T08:47:12] Andrew Rambaut:** https://www.nature.com/articles/s41467-020-19802-w/figures/2 [shared file(s): image.png] **[2021-07-02T08:47:36] Andrew Rambaut:** Maybe **[2021-07-02T08:48:21] Kristian G. Andersen:** Yeah, I think if we have the borders of those neighborhoods, then they could just be filled in **[2021-07-02T08:49:29] Andrew Rambaut:** I just mean trying to pick out those border on the little maps - some shapes that don't easily match **[2021-07-02T08:50:47] Kristian G. Andersen:** Yeah, I'm wondering if there's a vector of whatever map it is WHO used. Lemme dig a little.. **[2021-07-02T08:53:33] Andrew Rambaut:** https://els-jbs-prod-cdn.jbs.elsevierhealth.com/cms/attachment/5f763a8e-f8b4-4770-bab6-1f54fdcf8259/gr3.jpg [shared file(s): image.png] **[2021-07-02T08:53:49] Andrew Rambaut:** This one seems to fit a bit better **[2021-07-02T08:55:08] Kristian G. Andersen:** Somewhat - but still some inconsistency. **[2021-07-02T08:58:09] Kristian G. Andersen:** Friggin' impossible! **[2021-07-02T08:59:38] Andrew Rambaut:** Leave it - not worth it for the marginal point it would make. If we wanted to we could just cite the fact. **[2021-07-02T09:03:23] Andrew Rambaut:** Just say there was no excess deaths due to pneumonia or all-causes in JX in January? **[2021-07-02T09:10:45] Kristian G. Andersen:** I think the key point is that those excess deaths first cluster North of the river and then moves South - like the cases (cases have a sampling bias, excess deaths do not).We have a postdoc at our center who's from Wuhan - let me just try to ping him **[2021-07-02T09:13:58] Andrew Rambaut:** Not sure they show that: [shared file(s): image.png] **[2021-07-02T09:14:34] Andrew Rambaut:** Nothing in Week2, Week 3 has both sides of the river. **[2021-07-02T09:14:45] Andrew Rambaut:** But it is all Central Districts **[2021-07-02T09:15:07] Andrew Rambaut:** Spreading out to the outlying areas (which includes WIV) **[2021-07-02T09:17:27] Kristian G. Andersen:** There are two maps - the "pneumonia" one looks a little different [shared file(s): Screen Shot 2021-07-02 at 9.16.28AM.png] **[2021-07-02T09:18:03] Kristian G. Andersen:** (which I think is the one I have been going with). **[2021-07-02T09:18:38] Kristian G. Andersen:** Here's a better map where (I think) the districts are shown - I think some of the borders in the WHO map actuallycorrespond to rivers. [shared file(s): 20190322090211.png] **[2021-07-02T09:22:59] Andrew Rambaut:** Yes - that is better - you sent all-cause above (fig 20) but we want fig 21. **[2021-07-02T09:23:57] Kristian G. Andersen:** shared file(s): Map-of-the-administrative-divisions-of-Wuhan-city.png **[2021-07-02T09:24:14] Andrew Rambaut:** Again - we could just reference the WHO report and say that "by week 3, all pneumonia related excess mortality wasin districts to the north of the river" **[2021-07-02T09:25:00] Kristian G. Andersen:** Yeah, that might be easier - if it gets too busy, I think that's totally sufficient. **[2021-07-02T09:26:39] Andrew Rambaut:** I think week 3 is only: Dongxihu, Hanyang, Jiang'an, Jianghan **[2021-07-02T09:27:22] Andrew Rambaut:** By week 5 it is everywhere **[2021-07-02T09:28:14] Andrew Rambaut:** Perhaps we can highlight those 4 districts in the map and say they were the first to spike in excess mortality **[2021-07-02T09:29:13] Kristian G. Andersen:** Yeah, that'd be nice - and yes, I agree it's those four (although I have it as Dongxihu not Dongzihu) **[2021-07-02T09:29:56] Andrew Rambaut:** yes - mistyped... **[2021-07-02T09:33:14] Kristian G. Andersen:** And just as reference point - Huanan market is in Jianghan. **[2021-07-02T09:34:48] Andrew Rambaut:** Judging from Bob's map, Huanan market is in Jianghan? **[2021-07-02T09:36:01] Kristian G. Andersen:** Yeah - from Google too and just double-checked.