Spiracular

• Spiracular 15 Sep 2020 19:07 UTC

  9 points

  in reply to: Spiracular’s comment on: Coron­avirus Open Thread

  Chi­nese vi­rol­ogy re­searcher re­leased some­thing claiming it SARS-2 might even be ge­net­i­cally-ma­nipu­lated af­ter all? ZC45 and/​or ZXC21 back­bone. Claims that the RaTG13 genome was a con­cocted cover-up. I’m still as­sess­ing the sub-claims, but I think she makes some pretty co­gent-sound­ing ar­gu­ments for it.

  https://​​zen­odo.org/​​record/​​4028830#.X2EJo5NKj0v

  (That the RaTG13 se­quence could have been con­cocted is some­thing I kinda wish I’d thought of ear­lier.)

  See here for my fur­ther thoughts on this.

  EDIT: After as­sess­ing, I’m not find­ing the GMO claims con­vinc­ing. The RaTG13 story does seem to have some­thing weird go­ing on, and there’s sev­eral peo­ple and pa­pers that note weird in­con­sis­ten­cies (See the fur­ther thoughts, I don’t have a sim­ple ex­pla­na­tion.).

• Spiracular 15 Sep 2020 18:55 UTC

  2 points

  in reply to: jimrandomh’s comment on: Jim­ran­domh’s Short­form Posts

  Chi­nese vi­rol­ogy re­searcher re­leased some­thing claiming that SARS-2 might even be ge­net­i­cally-ma­nipu­lated af­ter all? After as­sess­ing, I’m not re­ally con­vinced of the GMO claims, but the RaTG13 story definitely seems to have some­thing weird go­ing on.

  Claims that the RaTG13 genome re­lease was a cover-up (it does look like some­thing’s fishy with RaTG13, al­though it might be differ­ent than Yan thinks). Claims ZC45 and/​or ZXC21 was the ac­tual back­bone (I’m feel­ing su­per-skep­ti­cal of this bit, but it has been hard for me to con­firm ei­ther way).

  https://​​zen­odo.org/​​record/​​4028830#.X2EJo5NKj0v (aka Yan Re­port)

  RaTG13 Looks Fishy

  Looks like some­thing fishy hap­pened with RaTG13, al­though I’m not con­vinced that ge­netic mod­ifi­ca­tion was in­volved. This is an ar­gu­ment built on pre-prints, but they ap­pear to offer sev­eral differ­ent lines of ev­i­dence that some­thing weird hap­pened here.

  Sim­plest story (via R&B): It looks like peo­ple first se­quenced this virus in 2016, un­der the name “BtCOV/​4991”, us­ing mine sam­ples from 2013. And for some rea­son, WIV re-re­leased the se­quence as “RaTG13″ at a later date?

  (edit: I may have just had a mi­s­un­der­stand­ing. Maybe BtCOV/​4991 is the name of the virus as se­quenced from miner-lungs, RaTG13 is the name of the virus as se­quenced from floor drop­pings? But in that case, why is the “fe­cal” sam­ple read­ing so weirdly low-bac­te­ria? And they prob­a­bly are em­bar­rassed that it took them that long to se­quence the fe­cal sam­ples, and should be.)

  A pa­per by by In­dian re­searchers Ra­halkar and Bahulikar ( https://​​doi.org/​​10.20944/​​preprints202005.0322.v1 ) notes that BtCoV/​​4991 se­quenced in 2016 by the same Wuhan Virol­ogy In­sti­tute re­searchers (and taken from 2013 sam­ples of a mi­ne­shaft that gave min­ers deadly pneu­mo­nia) was very similar, and likely the same, as RaTG13.

  A preprint by Ra­halkar and Bahulikar (R&B) ( doi: 10.20944/​preprints202008.0205.v1 ) notes that the frac­tion of bac­te­rial genomes in in the RaTG13 “fe­cal” sam­ple was ABSURDLY low (“only 0.7% in con­trast to 70-90% abun­dance in other fe­cal swabs from bats”). Some­thing’s weird there.

  A more re­cent weird dat­a­point: A pre-print Yan refer­enced ( https://​​www.ncbi.nlm.nih.gov/​​pmc/​​ar­ti­cles/​​PMC7337384/​​ ), whose find­ing (in graphs; it was left un­clear in their word­ing) was in­deed that a RaTG13 pro­tein didn’t com­pe­tently bind their Bat ACE2 sam­ples, but rather their Rat, Mouse, Hu­man, and Pig ACE2. It’s sup­pos­edly a horse­shoe bat virus (se­quenced by the Wuhan lab), so this seems hecka fishy to me.

  (Sure, their bat sam­ples weren’t pre­cisely the same species, but they tried 2 species from the same genus. SARS-2 DID bind for their R. macro­tis bat sam­ple, so it seems ex­tra-fishy to me that RaTG13 didn’t.).

  ((...oh. Ac­cord­ing to the R&B pa­per about the mi­ne­shaft, it was FILTY with rats, bats, poop, and fun­gus. And the CoV genome showed up in only one of ~280 sam­ples taken. If it’s like that, who the hell knew if it came from a rat or bat?))

  At this point, RaTG13 is gen­uinely look­ing pretty fishy to me. It might ac­tu­ally take ev­i­dence of a con­spir­acy the­ory in the other di­rec­tion for me to go back to neu­tral on that.

  E-Protein Similar­ity? Meh.

  I’m not find­ing the Protein-E se­quence similar­ity su­per-con­vinc­ing in it­self, be­cause while the logic is fine, it’s very mul­ti­ple-hy­poth­e­sis-test­ing fla­vored.

  I’m still look­ing into the ZC45 /​ ZXC21 claim, which I’m cur­rently feel­ing skep­ti­cal of. Here’s the pa­per that char­ac­ter­ized those: doi: 10.1038/​s41426-018-0155-5 . It’s true that it was by peo­ple work­ing at “Re­search In­sti­tute for Medicine of Nan­jing Com­mand.” How­ever, some­one on twit­ter used BLAST on the E-pro­tein se­quence, and found a gi­ant pile of differ­ent highly-re­lated SARS-like coro­n­aviruses. I’m try­ing to repli­cate that anal­y­sis us­ing BLAST my­self, and at a skim the 100% re­sults are all more SARS-CoV-2, and the close (95%) re­sults are damned di­verse. …I don’t see ZC in them, it looks like it wasn’t up­loaded. Ugh. (The E-pro­tein is only 75 amino acids long any­way. https://​​www.ncbi.nlm.nih.gov/​​pro­tein/​​QIH45055.1 )

  A differ­ent pa­per men­tions ex­treme S2-pro­tein similar­ity of early COVID-19 to ZC45 , but that pro­tein is highly-con­served. That makes this a less sur­pris­ing or mean­ingful re­sult. (E was claimed to be fast-evolv­ing, so its iden­ti­cal­ity would have been more sur­pris­ing, but I couldn’t con­firm it.) https://​​doi.org/​​10.1080/​​22221751.2020.1719902

  Other

  I think Yan offers a rea­son­able ar­gu­ment that a method could have been used that avoids ob­vi­ous ge­netic-mod­ifi­ca­tion “stitches,” in­stead us­ing meth­ods that are hard to dis­t­in­guish from nat­u­ral re­com­bi­na­tion events (ex: re­com­bi­na­tion in yeast). Sounds to­tally pos­si­ble to me.

  The fact that the early SARS-CoV-2 sam­ples were already quite adapted to hu­man ACE2 and didn’t have the rapid-evolu­tion you’d ex­pect from a fresh zoonotic in­fec­tion is some­thing a friend of mine had pre­vi­ously noted, prob­a­bly af­ter read­ing the fol­low­ing pa­per (recom­mended): https://​​www.biorxiv.org/​​con­tent/​​10.1101/​​2020.05.01.073262v1 (Zhan, Dev­er­man, Chan). This fact does seem fishy, and had already pushed me a bit to­wards the “Wuhan lab adap­ta­tion & es­cape” the­ory.

  What links here?

  • Spiracular's comment on March Coron­avirus Open Thread by Elizabeth (15 Sep 2020 19:07 UTC; 9 points)

• Spiracular 18 Aug 2020 3:17 UTC

  7 points

  on: Spirac­u­lar’s Short­form Feed

  On Com­mu­nity Co­or­di­na­tor Roles

  Some things I think help are:

  • Per­sonal fit to the par­tic­u­lar job de­mands (which can in­clude some sub­set of: lots of peo­ple-time, nav­i­gat­ing con­flict, fre­quent task-switch­ing, per­sonal ini­ti­a­tive, weigh­ing val­ues (or perform­ing triage), etc.)

  • Per­sonal in­vest­ment in ap­pro­pri­ate val­ues, and a last­ing com­mit­ment to do­ing the role well.

  • Hav­ing it be fairly clear what they’re bit­ing off, what re­sources they can use, and what ad­di­tional re­sources they can pe­ti­tion for. Hav­ing these things be ac­tu­ally fairly re­al­is­tic.

  (Min­i­mum ex­per­tise to thrive is prob­a­bly set by need­ing to find some­one who will ad­vo­cate for these things well enough to make them clear, or who can at least learn how to. Ne­bu­lous roles can be nav­i­gated, but it is harder, and in prac­tice means they need to be able to sur­vive re­defin­ing the role for them­selves sev­eral times.)

  • Hav­ing some con­nec­tions they (and oth­ers) re­spect who will help sup­port what (and how) they’re do­ing, lend some con­sis­tent sense of mean­ing to the work, and peo­ple who will offer re­li­able out­side judge­ment. Quite pos­si­bly, these things come from differ­ent peo­ple.

  (Public opinion is of­ten too noisy and fickle for most hu­man brains to learn off of it alone.)

  • For roles that re­quire the buy-in of oth­ers, it’s helpful if they have enough gen­eral re­spect or back­ing that they will mostly be treated as “le­gi­t­i­mate” in the role. But some roles re­ally only re­quire a small sub­set of high-buy-in peo­ple.

  • Also, some sort of step-down pro­ce­dure.

  Ex­tracted from a FB thread, where I was think­ing about burnout around neb­u­lously-defined com­mu­nity leader roles, and what pre­pared­ness and a good role would look like. Parts of this an­swer feel like I’m be­ing too vague and ob­vi­ous, but I thought it was worth mak­ing the list slightly more find­able.

• Spiracular 25 Jun 2020 23:52 UTC

  3 points

  in reply to: Kenny’s comment on: [META] Build­ing a whisper net­work.

  I’ve seen some pre­sen­ta­tions about how to do style-match­ing off of GitHub re­pos to pretty-con­fi­dently ID anony­mous coders. While set-up re­quires a siz­able amount of com­pute and data, the re­sults have got­ten quite im­pres­sive. There are ways to work against this (stuff that de­liber­ately ob­scures your cod­ing style, usu­ally by rewrit­ing your code), but they’re not that well known. And a similar thing can be done with writ­ing style and writ­ing sam­ples.

  Stay­ing anony­mous against high-effort at­tempts to dis­cern your iden­tity has got­ten very hard, and is only likely to get harder.

  At some point, all you can do is guard against the low-effort ones.

• Spiracular 25 Jun 2020 23:28 UTC

  4 points

  in reply to: siclabomines’s comment on: [META] Build­ing a whisper net­work.

  A quiz and a day’s wait be­fore adding a new user is an­other op­tion. Make it some­thing that a reg­u­lar lurker who read the rules would be able to pass eas­ily, but a rando couldn’t. SCP wiki did some­thing like this, it seemed to help with qual­ity con­trol.

  Ro­tate through 3 differ­ent quizes, or scram­ble the quiz or­der some­times, if you want to make au­to­mated sign-ups an­noy­ing for mobs and spam­mers. Have the web peo­ple track the num­ber of sign-up-quiz fails (it’s a nice met­ric for “is there a mob at the doorstep”).

  (Edit: Ah, some­one already pro­posed a more-elab­o­rate var­i­ant us­ing GPT-X. Sim­ple quizes with a few mild gotcha-ques­tions should be enough of a screen for most cases, though.)

  A pro­posal I think I haven’t seen posed is giv­ing new mem­bers a “trial pe­riod.” If an av­er­age (or ran­domly-se­lected) post doesn’t have a karma score of at least X by the end of the pe­riod (or if it dips be­low Y at any point), they’re out and their stuff is deleted. Ban them from hand­ing out karma un­til af­ter the trial, or this quickly breaks. This prob­a­bly still has weird in­cen­tive con­se­quences that I’m not see­ing, though...

  ...it does mean hav­ing a bit of an evap­o­ra­tive-filter for qual­ity-rat­ings, and it means links to crappy posts turn into dead­links in just a mat­ter of time.

• Spiracular 17 Jun 2020 22:53 UTC

  2 points

  in reply to: Kerry’s comment on: Does SARS-CoV-2 uti­lize an­ti­body-de­pen­dent en­hance­ment?

  I did spec­ify long-term, which for me meant time-frames of around a year to a decade out. Hon­estly, I sus­pect you’re largely right about the short-term.

  Well, ex­cept I might be more op­ti­mistic about vac­ci­na­tion efforts. Effec­tive vac­ci­na­tion pushes in the past give me some hope.

  Also, the mu­ta­tion rate is a good bit lower than the sea­sonal flu. SARS-CoV-2′s point-mu­ta­tions per year is around 28 sub­sti­tu­tions, which is about ¹⁄₂ as many as the flu. Or around ¹⁄₃ the rate, at ~1.1e-3 subs per site per year, com­pared to flu’s 3.3 subs per site per year. (Differ­ent genome lengths, hence the differ­ent an­swers.)

• Spiracular 11 Jun 2020 18:32 UTC

  2 points

  in reply to: siclabomines’s comment on: BBE W1: Per­sonal Note­tak­ing Desider­ata Walkthrough

  It’s kinda weird to me how limited the op­tions seem to be for flash­cards and an­no­ta­tion/​margina­lia. While plenty of things perform the core func­tion­al­ity, I haven’t seen any­where near as many in­ter­est­ing ex­per­i­ments with these for­mats as I have with out­lin­ers.

  And for flash­cards par­tic­u­larly? You’d think it’d be the sim­plest damn thing to pro­gram!

  I’m not much of an in­dex card per­son, but there seem to be a lot of peo­ple who swear on elab­o­rate in­dex-card set-ups. That I haven’t seen more of them im­ple­mented as soft­ware con­fuses me too.

  (It does seem like it would be a good use of flash­cards to set up a rudi­men­tary pri­or­ity queue, or stack. Pos­si­bly with added ran­dom­ness. I’m hon­estly a lit­tle sur­prised it doesn’t al­low that.)

• Spiracular 11 Jun 2020 18:10 UTC

  1 point

  in reply to: quanticle’s comment on: BBE W1: HMCM and Note­tak­ing Systems

  I think this is a good ques­tion. Here are some prob­a­ble com­po­nents of pro­gramma­bil­ity...

  • Did it sur­face most of its ac­tual func­tion­al­ity to users?

    • A cou­ple differ­ent set­tings: Closed pro­pri­etary cloud soft­ware, API (how friendly or per­mis­sive is it?), down­load­able open-source...

  • How easy (and safe!) it is to call rele­vant util­ity func­tions?

    • Do you need to close the soft­ware to edit it? Did they merely sur­face the func­tion­al­ity, or did they also leave func­tions that were highly-ex­posed, la­beled, well-doc­u­mented, and easy to use? How well do they ad­here to var­i­ous stan­dards, and there­fore benefit from skill-trans­fer? Is it easy to screw up? To re­vert? What’s the learn­ing curve like?

• Spiracular 11 Jun 2020 17:51 UTC

  1 point

  on: BBE W1: HMCM and Note­tak­ing Systems

  TextCards

  I re­ally wish there was bet­ter flash­card and an­no­ta­tion/​margina­lia soft­ware out there! It’s kinda weird to me how limited the op­tions seem to be for both. While plenty of things perform the core func­tion­al­ity, I haven’t seen as many in­ter­est­ing ex­per­i­ments with it as I have with, say, out­lin­ers.

  While writ­ing this post, I de­vel­oped a vague sus­pi­cion that there’s some­thing in-be­tween An­no­ta­tor and Flash­card that could be pretty valuable if some­one ac­tu­ally im­ple­mented it. This seems as good a place as any to de­scribe it. (And if some­one has already done it, or wants to do it, cool!)

  An­no­ta­tors and Flash­cards are both of­ten track­ing an un­der­ly­ing dic­tio­nary-ish data-type, and a lot of flash­cards seem to origi­nate from text­books. I have a sus­pi­cion that there should ex­ist a good stan­dard­ized-for­mat note­taker that goes some­thing like… this?

  Tex­tCards: 3 linked items

  • A bounded sec­tion of high­lighted text­book (Any size, from a sec­tion to en­tire chap­ter. Some­times dis­con­tin­u­ous.)

  • An in­dex-card la­conic de­scrip­tion (or an­swer)

  • A ti­tle (or ques­tion)

  Some­times, it could be used to pose stan­dard quiz-ques­tions (the high­lighted sec­tion is just the part of the book the quiz came from, the ti­tle is the ques­tion, the de­scrip­tion is an an­swer). But where it might re­ally shine is in “Sum­ma­rize Chap­ter X” ques­tions; it en­courages you to write along as you read the text, and if you miss some­thing on a quiz, you can click right to the sec­tions you were origi­nally sum­ma­riz­ing.

  When ren­dered as margina­lia, the small ti­tles (un­til click) should make that ex­pe­rience more tol­er­able for fre­quent-mar­gin-users. (Margina­lia async­ing with the page seems like a re­ally com­mon prob­lem, oth­er­wise.)

  For con­ve­nience, adding some­thing that swipes all of the ques­tions from a high­lighted sec­tion of the text to form the front end of flash­cards (that you then an­swer) seems pretty nice. For well-for­mat­ted an­swer-sec­tions, you might even be able to get it to pair the two (but you’d prob­a­bly need to high­light where to look). Ad­di­tion­ally, it wouldn’t be that hard for it to track which chap­ter’s ques­tions you’re do­ing poorly on -and there­fore what chap­ters you should re-read- if it knows where in the book you swiped them from. Bonus points if you can sort and cross-link notes by ti­tle, folder, tags, over­lap­ping high­lights, and/​or or­der in text.

  Pre­sum­ably this is usu­ally harder than I think it should be, be­cause PDFs are just awful (I’ve dragged ta­bles from PDFs be­fore; I feel so sorry for Tab­ula!). But HTML books and ebooks don’t have that prob­lem, and of­ten simu­late a text­book-like struc­ture.

BBE W1: HMCM and Note­tak­ing Systems

BBE W1: Per­sonal Note­tak­ing Desider­ata Walkthrough

Build a Bet­ter Exo­brain: Week 0 Overview

• Spiracular 8 Jun 2020 22:22 UTC

  1 point

  in reply to: mr-hire’s comment on: Spirac­u­lar’s Short­form Feed

  At min­i­mum it’s over­sim­plified. That’s why I called it a parable.

  I ap­pre­ci­ate the nu­ance in your com­ment even as it is. But I’m cu­ri­ous about the facts/​nar­ra­tives you have that dis­agree with it?

  (My per­sonal strongest mur­phyjitsu was that “space pro­gram de­fund­ing” was more com­pli­cated than this. If you know more about that than I do, I’d be cu­ri­ous to hear about it.)

• Spiracular 4 Jun 2020 18:43 UTC

  15 points

  on: Spirac­u­lar’s Short­form Feed

  A Parable on Vi­sual Impact

  A long time ago, you could get the biggest pos­i­tive vi­sual im­pact for your money by gen­er­at­ing art, and if you wanted awe you could fund gar­dens and cathe­drals. And lo, these ar­eas were well-funded!

  The print­ing press ar­rived. Now, you could get mas­sive num­bers of pam­phlets and wood­cuts for a frac­tion of the price of paint­ings. And lo, these ar­eas were well-funded!

  Then tv ap­peared. Now, if you wanted the great­est awe and the biggest pos­i­tive vi­sual im­pact for your money, you crafted some­thing suit­able for the new medium. And in that time, there were mas­sive made-for-tv pro­pa­ganda cam­paigns, and money poured into de­vel­op­ing space­craft, and we got our first awe-in­spiring images of Earth from the moon.

  Some even claim the Soviet Union was defeated by the view through a tele­vi­sion screen of a bet­ter life in Amer­ica.

  And then we de­vel­oped CGI and 3D MMORPGs. And lo, the space pro­gram de­funded, as peo­ple built en­tire cities, en­tire planets in CGI for a tiny frac­tion of the cost!

• Spiracular 12 May 2020 23:49 UTC

  12 points

  in reply to: TurnTrout’s comment on: Stop say­ing wrong things

  I didn’t down­vote it, but I did find the writ­ing style mildly grat­ing. (Re­lat­edly: I can­not ex­ceed your set-point of smug, it is over 9000 :) )

  To be clear: I think the un­der­ly­ing point was pretty good, and I mostly had is­sues with the de­liv­ery. I still feel it was prob­a­bly some­thing worth writ­ing, al­though I also think I’m not the tar­get au­di­ence for this par­tic­u­lar bit of ad­vice.

  Some of it was prob­a­bly a tone thing, which I won’t go into. But here are some things that seem tractable:

  My ex­pe­rience of it was a bit bet­ter as soon as I switched out al­most all the “You”s for “I”s. I have some­thing of a dis­taste for the… pup­pety-feel­ing where some­one seems to be try­ing to put words into my mouth, that don’t fit with my ac­tual ex­pe­rience. This set it off pretty badly. There are a lot of speci­fics, and it’s clearly your per­sonal ac­count; own it.*

  This got re­ally stark for me at around...

  First, you just end up be­ing an ass­hole pre­tend­ing to be honest

  ...which in­stantly broke my im­mer­sion. My ex­pe­rience of be­ing painfully hon­est with my­self, and then oth­ers, was rad­i­cally differ­ent.**

  It prob­a­bly also could have used more short para­graphs, and some va­ri­ety in pre­sen­ta­tion. Some of the goals you’ve acted on and then for­got­ten, such as Real Anal­y­sis or Man­darin, could have been bet­ter-pre­sented as sin­gle bul­let-points af­ter go­ing into only 1 of them in-depth. The Etsy sec­tion could have used a header, and been bro­ken into more than 1 para­graph. That kind of thing.

  *A lot of bad ad­vice on per­sua­sive es­say writ­ing en­courages the for­ma­tion of habits like this. One can­not list the num­ber of times one has been told to make that un­nat­u­ral sub­sti­tu­tion of “One,” where “I” would have been bet­ter, and more hon­est. Teach­ers who do this are just… wrong. Tech­ni­cal writ­ing is a real thing, but this way of teach­ing it is crap, and can ruin oth­er­wise-de­cent writ­ers.

  ** My ex­pe­rience was close to painful self-con­scious­ness (for self-hon­esty), and weird so­cial penalties (for hon­esty with oth­ers). Real hon­esty is of­ten dis­tinctly un-charm­ing, but in my case… blunt­ness leaned closer to “overly-in­vested*** ec­cen­tric” than “ass­hole.” If it had been framed as a self-ac­count, this jar­ring wouldn’t have been an is­sue.

  *** Ex­hibit A: This over­grown write-up.

• Spiracular 12 May 2020 21:48 UTC

  5 points

  on: Why don’t we tape sur­gi­cal masks to the face to seal them air­tight?

  I’ve per­son­ally found that just sta­pling shut the outer edge of the first fold on ei­ther side of a sur­gi­cal mask re­sults in a mask that mostly sucks to my face when I breathe in. It doesn’t stick well when I breathe out, though.

  This is much eas­ier to im­ple­ment than fancy ad­her­ents; all it takes is 2 sta­ples and a sta­pler, or a nee­dle and thread. It struck me as a plau­si­ble 80-20. (Well, less than 80%. Ob­vi­ously, this is no N95.)

  (It might only work for some face-shapes, though.)

  This gen­er­ates a new prob­lem, which would also ap­ply to taped/​glued solu­tions:

  The front is made of a softer fabric than N95s. After an hour of wear, it will suck to my lips if I breathe in with my mouth. And since it doesn’t stick on out-breaths, air still gets out the sides and some­times the top. I don’t know how much ad­di­tional risk this pre­sents, but I would be cu­ri­ous to hear some­one weigh in.

  (It also has un­com­fortably-high hu­midity, but that’s even more true of N95s.)

• Spiracular 9 May 2020 19:32 UTC

  3 points

  on: April Coron­avirus Open Thread

  Late Edit: Pan­golins with this viral in­fec­tion have been found from both Guangxi and Guang­dong provinces.

  De­spite the virus be­ing char­ac­ter­ized in pan­golins, af­ter look­ing into this, I now think it is ba­si­cally in­cor­rect to think of this as pri­mar­ily a “pan­golin virus.” The pan­golins were a dy­ing ca­nary in a coal mine, and prob­a­bly caught it from some­thing else that serves as the real reser­voir species for this nCOV pre­cur­sor*.

  Th­ese pan­golins were be­ing smug­gled when they were cap­tured by the au­thor­i­ties in Guangxi. They were dy­ing of prob­a­bly sev­eral dis­eases; they had le­sions in their skin, in­tense con­ges­tion, and were in gen­er­ally atro­cious con­di­tion when they got se­quenced for viruses. They turned up pos­i­tive for all man­ner of things (her­pes out the wa­zoo, but also a sendai virus which was most closely re­lated to the se­quence of a hu­man-taken sam­ple, a paramyx­ovirus, and yes, sev­eral coro­n­aviruses).

  Here’s the origi­nal ar­ti­cle on the pan­golins whose vi­rome they se­quenced, and the ar­ti­cle not­ing its re­lat­ed­ness to nCOV.

  Given that so many of the pan­golins died, the pan­golins look more like a highly-sus­cep­ti­ble sec­ondary species, than a mostly-asymp­tomatic pri­mary re­serve species* to me.

  * GD/​PIL or GD/​P2S is thought of as a pos­si­ble nCOV pro­gen­i­tor, alongside bat-virus RaTG13. GD/​PIL’s re­cep­tor-bind­ing mo­tif (RBM) in par­tic­u­lar is iden­ti­cal to SARS-2′s, al­though nCOV oth­er­wise ap­pears more closely re­lated to RaTG13.

  ** On ed­u­cated pri­ors, I think the true reser­voir is prob­a­bly rats, bats, or (less likely) hu­mans in the Guanxi, Hu­nan, and/​or Hubei province.

  Per­son­ally, I as­sign >90% on ei­ther rats (strong pri­ors + skin le­sions) or bats (strong pri­ors + sim­plest story). But these were ex­otic an­i­mal smug­glers; there is a small chance that the origi­nal reser­voir species could be any an­i­mal.

  What links here?

  • Spiracular's comment on Jim­ran­domh’s Shortform by jimrandomh (9 May 2020 19:28 UTC; 4 points)
  • Spiracular's comment on March Coron­avirus Open Thread by Elizabeth (1 May 2020 2:07 UTC; 1 point)

• Spiracular 9 May 2020 19:28 UTC

  4 points

  in reply to: Lukas_Gloor’s comment on: Jim­ran­domh’s Short­form Posts

  I think it’s prob­a­bly a virus that was merely iden­ti­fied in pan­golins, but whose pri­mary host is prob­a­bly not pan­golins.

  The pan­golins they se­quenced weren’t asymp­tomatic car­ri­ers at all; they were sad smug­gled spec­i­mens that were dy­ing of many differ­ent dis­eases si­mul­ta­neously.

  I looked into this semi-re­cently, and wrote up some­thing here.

  ---

  The pan­golins were ap­pre­hended in Guangxi, which shares some of its bor­der with Yun­nan. Nei­ther of these provinces are di­rectly con­tigu­ous with Hubei (Wuhan’s province), fwiw. (map)

• Spiracular 7 May 2020 15:45 UTC

  2 points

  in reply to: Mary Chernyshenko’s comment on: Cop­ing and Cultures

  True, and I’ve seen lab work cul­ti­vate some­thing similar.

  (I’m pretty sure this par­tic­u­lar skill is the in­verse of pro­gram­mer-style “laz­i­ness,” fun­nily enough. In one field, see­ing rep­e­ti­tion is re­as­sur­ing. In the other, it can be ev­i­dence that your code is not as el­e­gant and mod­u­larized as it could be.)

  I always thought you’d au­to­mat­i­cally learn the gait if you just did the work of­ten enough, though. It’s definitely a cop­ing skill, but I read its ori­gins as more cul­ti­vated than cul­turally-in­duced or taught.

  It mostly fol­lows the nat­u­ral in­cen­tive gra­di­ents of the work. This can be in con­trast to things like sep­a­ra­tion of self and client in psy­chol­ogy, which seems to feel ac­tively un-nat­u­ral for many peo­ple. Of course, there’s some­thing of a spec­trum here, with heavy in­di­vi­d­ual vari­a­tion.

• Spiracular 7 May 2020 6:28 UTC

  1 point

  in reply to: Ruby’s comment on: Cop­ing and Cultures

  Thanks! Sounds like a promis­ing lead.