Hibernation

[Nickleaton]

His next claim is that it is very unlikely that there would be bats naturally living in the metropolitan distict of Wuhan, and in fact no bats were traded at the market at all.

Let us go one step further: there were no bats in Wuhan in December, because bats hibernate in the winter!

=========

I agree - unlikely to be traded. The hibernation applies to bats overflying, dropping infected faeces. Again very unlikely.

However if traded, its easier to catch a hibernating bat than an active one.

[angoffinet]

I agree with all that, just that people in conflict of interest refuse to hear anything, even when substantiated with solid arguments like you do. If I cound twitter users, project-evidence followers etc..., I am not sure we reach 100 to support actively the leak scenario. From what I hear around, many people believe a leak occurred and some even think it is widely accepted. But this is far from being the case with most professional virologists and epidemiologists promoting the "natural" "zoonosis" origin. Quite discouraging, but what can we do apart from keeping our modest pressure? Take care, Andre

Andre GOFFINET, MD, PhD Prof. em. Institute of Neuroscience University of Louvain, Belgium +32 (0) 473 899818

Le lun. 6 juil. 2020 à 08:28, Gilles Demaneuf notifications@github.com a écrit :

I agree that it is odd. But to be fair if one believes in Bayesian analysis then one should still give it a go. The beauty about this anyway is whatever range you give to each probability, there is no way that the US virus hypothesis even register against the lab accident or the zoonotic event. If someone tells me that they still believe in it, then I just ask them to give me their probability estimates that are consistent with their conclusion. They can't so it - as simple as that.

The same goes with the zoonotic hypothesis. The fact that the outbreak started in Wuhan, and only in Wuhan, knock that one back about 2 orders against the lab accident. Both the 4 previous SARS lab release in Chinese lab and the warning (should I call it red-flag waving?) from Yuan Zhiming show that lab accident probabilities are at least (to be very generous) as likely in Wuhan than in the rest of the world. If you consider the number of viruses, projects, labs involved in Wuhan then it becomes insolently high and risky.

This is my second conclusion Andre: not only probabilities show us that most likely covid-19 was a lab accident, but the probability of another lab accident of the same sort is over the next 10 years is just as high, if not higher as more P3 and P4 are being built.

We need to get this message out, backed by a probability analysis. This is endangering us all.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-654041188, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEX3H6LY7STFA3XPE7GAKRLR2FVKLANCNFSM4MYUF6PQ .

[research-project-cov]

WHO reports on P4 & P3+ labs in the world , dated 2018; at that time 48 operational or newly constructed, 3 lab under construction, 3 planned; maybe more planned now; In Wuhan, "The building features 3000 m2 of BSL-4 space, including four independent laboratories areas and two animal suites, in addition to 20 BSL-2 and two BSL-3 laboratories. " in the world there are thousands of BSL-3; although BSL-4 is to bring more safety, has any one calculated the risk on a global point of view, with feasible causes and likely consequences (more lab, more risk?) WHO-WHE-CPI-2018.40-eng.pdf

[angoffinet]

Currently, China has at least 3 P4.

1. Wuhan WIV. 2 Harbin veterinary institute of CAS, animal BSL4, see paper in Science https://science.sciencemag.org/content/368/6494/1016 BTW, I suspect that a leak during this crasy bunch of infection in february described in that paper contributed to the mini-outbreak in Harbin in april. 3; Kunming, see bioRxiv paper: https://www.biorxiv.org/content/10.1101/2020.04.08.031807v1 At least two others are planned, in Beijing and Guanzhou. If nobody reacts, SARS3 will follow in a few years.

Andre GOFFINET, MD, PhD Prof. em. Institute of Neuroscience University of Louvain, Belgium +32 (0) 473 899818

Le lun. 6 juil. 2020 à 09:22, anonymous notifications@github.com a écrit :

WHO reports on P4 & P3+ labs in the world , dated 2018; at that time 48 operational or newly constructed, 3 lab under construction, 3 planned; maybe more planned now; In Wuhan, "The building features 3000 m2 of BSL-4 space, including four independent laboratories areas and two animal suites, in addition to 20 BSL-2 and two BSL-3 laboratories. " in the world there are thousands of BSL-3; although BSL-4 is to bring more safety, has any one calculated the risk on a global point of view, with feasible causes and likely consequences (more lab, more risk?) WHO-WHE-CPI-2018.40-eng.pdf https://github.com/Project-Evidence/project-evidence.github.io/files/4877197/WHO-WHE-CPI-2018.40-eng.pdf

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-654061871, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEX3H6LGZDQUI2ARXGRZMMTR2F3R7ANCNFSM4MYUF6PQ .

[Nickleaton]

Wich gives 1/160 of 1 event in 10 year chance of an accidental accident in a US lab, with an accidental contamination chain from that lab accident to the Wuhan known cases, going through the Wuhan games. =========== =========== How many leaks have we had from US labs to date over how many years? I bet you that it is way more than one.

On Mon, 6 Jul 2020 at 02:35, Gilles Demaneuf notifications@github.com wrote:

Well, we can always try to bracket the probability of even the wildest theory, so let's give it a go. Firs t we need to clarify the hypothesis here, which is not very easy as the Chinese narative changed quite a lot. I believe that it can be summarized as 'Covid19 was brought to Wuhan by US participants during the Wuhan military games'. OK, let's go with that.

Let's look at the story origin and the supposed source and chain of contamination.

-

First there is a big problem with the chain: China has never even hinted to a known link between a Wuhan infection and someone attending the Games. No that's surprising. You would suspect that by now they would have been able to find a chain of contamination going back there. But no. None. It's a black hole - there is no known chain. That in itself knocks the theory back by a few orders. (The opposite would be true: if the early chains of infections were pointing to the Wuhan game then the probability would be high that the hypothesis is valid).

Then there are the contradictions in the story. We were told that some athletes were treated for covid like symptoms in an hospital in China, when the same hospital later clarified they were treated for malaria. Then we were given the name of a US athlete (which was first floated by a US based conspiracy theorist). Again all hot air - that athlete did not have covid. Does not make it very realistic.

Then we can look at the supposed source. There there are two alternatives: accidental contamination in the US of a/some participant(s) and malevolent contamination.

1. Accidental contamination. The problem is that there are no Souther Chinese bat populations in the US as you can imagine. So it would have to be accidental contamination at a lab. Great. But wait a second, between the probability of an accidental contamination in a US lab and the probability of an accidental contamination in a Wuhan lab, which one do you think is the highest given the security infrastructure and the number of covid projects worked on? I would say the probability of a covid lab accident in the US (if not the whole rest of the world) is at most 1/4th the probability of the same lab accident in Wuhan. But that's not all. Because we are talking about a US covid 19 accident with no known cases in the US and only knows cases in Wuhan. So you could look at something like the total population of the major cities around US lab against the Wuhan population and also multiply that ratio with something like 1/10 for the chance that a lab employee (or someone close to an asymptomatic lab employee) went to Wuhan as part of the US delegation. At the end of the day you will get something like 1/4 x 1/2 x 1/10 = 1/80. Other alternative: malevolent contamination. Sorry but that is pure conspiracy theory, and it makes no sense whatsoever given the US lack of preparation.

So putting things together:

• source: x1/80 * 1 event every 10 years
• chain: maybe x1/2 (due to its weakness - no known)
• story fuzziness: I will be generous and not reduce the probabilities
• so x1

Wich gives 1/160 of 1 event in 10 year chance of an accidental accident in a US lab, with an accidental contamination chain from that lab accident to the Wuhan known cases, going through the Wuhan games.

Basically it does not even register against the reasonable 1 event in 10 year chance of a coronavirus lab accident in Wuhan.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-653972427, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAF7NEBFCPPZCT4LWYRAA6DR2ES5LANCNFSM4MYUF6PQ .

-- Nick

[Nickleaton]

Claiming it originated outside China is pretty pathetic.

You have to look at the motivation and who is the message being targeted at. In this case its being targeted at the domestic Chinese audience. The CCP cannot admit its a local leak.

On Mon, 6 Jul 2020 at 06:49, Andre Goffinet notifications@github.com wrote:

Claiming it originated outside China is pretty pathetic. They blamed USA and military games, sounding quite ridiculous to say the least. A few weeks ago, they hinted that the origin may be in Italy, where bats (pipistrella) have been hosted in cities to fight mosquitoes (which is true, but no coronavirus in those bats as far as we know). The chinese narrative is incredible all along and reads more like panic-driven than anything plausible. Not sure that bayesian analysis can help much when dealing with low odds like this. But it is worth trying, and an elegant and useful case study.

Andre GOFFINET, MD, PhD Prof. em. Institute of Neuroscience University of Louvain, Belgium +32 (0) 473 899818

Le lun. 6 juil. 2020 à 03:35, Gilles Demaneuf notifications@github.com a écrit :

Well, we can always try to bracket the probability of even the wildest theory, so let's give it a go. Firs t we need to clarify the hypothesis here, which is not very easy as the Chinese narative changed quite a lot. I believe that it can be summarized as 'Covid19 was brought to Wuhan by US participants during the Wuhan military games'. OK, let's go with that.

Let's look at the story origin and the supposed source and chain of contamination.

-

First there is a big problem with the chain: China has never even hinted to a known link between a Wuhan infection and someone attending the Games. No that's surprising. You would suspect that by now they would have been able to find a chain of contamination going back there. But no. None. It's a black hole - there is no known chain. That in itself knocks the theory back by a few orders. (The opposite would be true: if the early chains of infections were pointing to the Wuhan game then the probability would be high that the hypothesis is valid).

Then there are the contradictions in the story. We were told that some athletes were treated for covid like symptoms in an hospital in China, when the same hospital later clarified they were treated for malaria. Then we were given the name of a US athlete (which was first floated by a US based conspiracy theorist). Again all hot air - that athlete did not have covid. Does not make it very realistic.

Then we can look at the supposed source. There there are two alternatives: accidental contamination in the US of a/some participant(s) and malevolent contamination.

1. Accidental contamination. The problem is that there are no Souther Chinese bat populations in the US as you can imagine. So it would have to be accidental contamination at a lab. Great. But wait a second, between the probability of an accidental contamination in a US lab and the probability of an accidental contamination in a Wuhan lab, which one do you think is the highest given the security infrastructure and the number of covid projects worked on? I would say the probability of a covid lab accident in the US (if not the whole rest of the world) is at most 1/4th the probability of the same lab accident in Wuhan. But that's not all. Because we are talking about a US covid 19 accident with no known cases in the US and only knows cases in Wuhan. So you could look at something like the total population of the major cities around US lab against the Wuhan population and also multiply that ratio with something like 1/10 for the chance that a lab employee (or someone close to an asymptomatic lab employee) went to Wuhan as part of the US delegation. At the end of the day you will get something like 1/4 x 1/2 x 1/10 = 1/80. Other alternative: malevolent contamination. Sorry but that is pure conspiracy theory, and it makes no sense whatsoever given the US lack of preparation.

So putting things together:

• source: x1/80 * 1 event every 10 years
• chain: maybe x1/2 (due to its weakness - no known)
• story fuzziness: I will be generous and not reduce the probabilities
• so x1

Wich gives 1/160 of 1 event in 10 year chance of an accidental accident in a US lab, with an accidental contamination chain from that lab accident to the Wuhan known cases, going through the Wuhan games.

Basically it does not even register against the reasonable 1 event in 10 year chance of a coronavirus lab accident in Wuhan.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub < https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-653972427 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/AEX3H6KHAOZBBZ4JP3JGJBLR2ES5LANCNFSM4MYUF6PQ

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-654028242, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAF7NEDGT7GU5EA7CMRADCLR2FQUXANCNFSM4MYUF6PQ .

-- Nick

[angoffinet]

Leaks from US labs are documented in project-evidence. There were really very few. In P4, one leak is one too many.

Andre GOFFINET, MD, PhD Prof. em. Institute of Neuroscience University of Louvain, Belgium +32 (0) 473 899818

Le lun. 6 juil. 2020 à 10:09, Nickleaton notifications@github.com a écrit :

Wich gives 1/160 of 1 event in 10 year chance of an accidental accident in a US lab, with an accidental contamination chain from that lab accident to the Wuhan known cases, going through the Wuhan games. =========== =========== How many leaks have we had from US labs to date over how many years? I bet you that it is way more than one.

On Mon, 6 Jul 2020 at 02:35, Gilles Demaneuf notifications@github.com wrote:

Well, we can always try to bracket the probability of even the wildest theory, so let's give it a go. Firs t we need to clarify the hypothesis here, which is not very easy as the Chinese narative changed quite a lot. I believe that it can be summarized as 'Covid19 was brought to Wuhan by US participants during the Wuhan military games'. OK, let's go with that.

Let's look at the story origin and the supposed source and chain of contamination.

-

First there is a big problem with the chain: China has never even hinted to a known link between a Wuhan infection and someone attending the Games. No that's surprising. You would suspect that by now they would have been able to find a chain of contamination going back there. But no. None. It's a black hole - there is no known chain. That in itself knocks the theory back by a few orders. (The opposite would be true: if the early chains of infections were pointing to the Wuhan game then the probability would be high that the hypothesis is valid).

Then there are the contradictions in the story. We were told that some athletes were treated for covid like symptoms in an hospital in China, when the same hospital later clarified they were treated for malaria. Then we were given the name of a US athlete (which was first floated by a US based conspiracy theorist). Again all hot air - that athlete did not have covid. Does not make it very realistic.

Then we can look at the supposed source. There there are two alternatives: accidental contamination in the US of a/some participant(s) and malevolent contamination.

1. Accidental contamination. The problem is that there are no Souther Chinese bat populations in the US as you can imagine. So it would have to be accidental contamination at a lab. Great. But wait a second, between the probability of an accidental contamination in a US lab and the probability of an accidental contamination in a Wuhan lab, which one do you think is the highest given the security infrastructure and the number of covid projects worked on? I would say the probability of a covid lab accident in the US (if not the whole rest of the world) is at most 1/4th the probability of the same lab accident in Wuhan. But that's not all. Because we are talking about a US covid 19 accident with no known cases in the US and only knows cases in Wuhan. So you could look at something like the total population of the major cities around US lab against the Wuhan population and also multiply that ratio with something like 1/10 for the chance that a lab employee (or someone close to an asymptomatic lab employee) went to Wuhan as part of the US delegation. At the end of the day you will get something like 1/4 x 1/2 x 1/10 = 1/80. Other alternative: malevolent contamination. Sorry but that is pure conspiracy theory, and it makes no sense whatsoever given the US lack of preparation.

So putting things together:

• source: x1/80 * 1 event every 10 years
• chain: maybe x1/2 (due to its weakness - no known)
• story fuzziness: I will be generous and not reduce the probabilities
• so x1

Wich gives 1/160 of 1 event in 10 year chance of an accidental accident in a US lab, with an accidental contamination chain from that lab accident to the Wuhan known cases, going through the Wuhan games.

Basically it does not even register against the reasonable 1 event in 10 year chance of a coronavirus lab accident in Wuhan.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub < https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-653972427 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/AAF7NEBFCPPZCT4LWYRAA6DR2ES5LANCNFSM4MYUF6PQ

.

-- Nick

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-654084232, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEX3H6PWHLGXAAXSJ7EM66LR2GBDPANCNFSM4MYUF6PQ .

[Nickleaton]

Few. If the odds for the US are 1 in 140 years and there are a few, then the conclusion has to be that the 1 in 140 year odds are wrong. Simple science. Theory, predict test. If the test fails, the theory is wrong. If the test passes, then you have more evidence its correct. You don't have proof its correct.

On Mon, 6 Jul 2020 at 10:46, Andre Goffinet notifications@github.com wrote:

Leaks from US labs are documented in project-evidence. There were really very few. In P4, one leak is one too many.

Andre GOFFINET, MD, PhD Prof. em. Institute of Neuroscience University of Louvain, Belgium +32 (0) 473 899818

Le lun. 6 juil. 2020 à 10:09, Nickleaton notifications@github.com a écrit :

Wich gives 1/160 of 1 event in 10 year chance of an accidental accident in a US lab, with an accidental contamination chain from that lab accident to the Wuhan known cases, going through the Wuhan games. =========== =========== How many leaks have we had from US labs to date over how many years? I bet you that it is way more than one.

On Mon, 6 Jul 2020 at 02:35, Gilles Demaneuf notifications@github.com wrote:

Well, we can always try to bracket the probability of even the wildest theory, so let's give it a go. Firs t we need to clarify the hypothesis here, which is not very easy as the Chinese narative changed quite a lot. I believe that it can be summarized as 'Covid19 was brought to Wuhan by US participants during the Wuhan military games'. OK, let's go with that.

Let's look at the story origin and the supposed source and chain of contamination.

-

First there is a big problem with the chain: China has never even hinted to a known link between a Wuhan infection and someone attending the Games. No that's surprising. You would suspect that by now they would have been able to find a chain of contamination going back there. But no. None. It's a black hole - there is no known chain. That in itself knocks the theory back by a few orders. (The opposite would be true: if the early chains of infections were pointing to the Wuhan game then the probability would be high that the hypothesis is valid).

Then there are the contradictions in the story. We were told that some athletes were treated for covid like symptoms in an hospital in China, when the same hospital later clarified they were treated for malaria. Then we were given the name of a US athlete (which was first floated by a US based conspiracy theorist). Again all hot air - that athlete did not have covid. Does not make it very realistic.

Then we can look at the supposed source. There there are two alternatives: accidental contamination in the US of a/some participant(s) and malevolent contamination.

1. Accidental contamination. The problem is that there are no Souther Chinese bat populations in the US as you can imagine. So it would have to be accidental contamination at a lab. Great. But wait a second, between the probability of an accidental contamination in a US lab and the probability of an accidental contamination in a Wuhan lab, which one do you think is the highest given the security infrastructure and the number of covid projects worked on? I would say the probability of a covid lab accident in the US (if not the whole rest of the world) is at most 1/4th the probability of the same lab accident in Wuhan. But that's not all. Because we are talking about a US covid 19 accident with no known cases in the US and only knows cases in Wuhan. So you could look at something like the total population of the major cities around US lab against the Wuhan population and also multiply that ratio with something like 1/10 for the chance that a lab employee (or someone close to an asymptomatic lab employee) went to Wuhan as part of the US delegation. At the end of the day you will get something like 1/4 x 1/2 x 1/10 = 1/80. Other alternative: malevolent contamination. Sorry but that is pure conspiracy theory, and it makes no sense whatsoever given the US lack of preparation.

So putting things together:

• source: x1/80 * 1 event every 10 years
• chain: maybe x1/2 (due to its weakness - no known)
• story fuzziness: I will be generous and not reduce the probabilities
• so x1

Wich gives 1/160 of 1 event in 10 year chance of an accidental accident in a US lab, with an accidental contamination chain from that lab accident to the Wuhan known cases, going through the Wuhan games.

Basically it does not even register against the reasonable 1 event in 10 year chance of a coronavirus lab accident in Wuhan.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <

https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-653972427

, or unsubscribe <

https://github.com/notifications/unsubscribe-auth/AAF7NEBFCPPZCT4LWYRAA6DR2ES5LANCNFSM4MYUF6PQ

.

-- Nick

— You are receiving this because you commented. Reply to this email directly, view it on GitHub < https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-654084232 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/AEX3H6PWHLGXAAXSJ7EM66LR2GBDPANCNFSM4MYUF6PQ

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-654131422, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAF7NEEGQGAWLAUGCJBVBF3R2GMOZANCNFSM4MYUF6PQ .

-- Nick

[nemonominem]

Nick, we need to look at the leaks of coronaviruses specifically. If there are 10 P4 labs in the US but none is studying coronaviruses it does add to any probability of a coronavirus leak from these labs.

Also you should look at P2 to P4 labs in China. Most coronaviruses were studied at low P2 in Wuhan. SARS is a coronavirus and there were 4 leaks recently in China - that gives you a good idea of the calibration for Chinese labs studying coronaviruses. Also the article by Yuan Zhiming confirms that there risk in Chinese labs is very much present (https://www.sciencedirect.com/science/article/pii/S2588933819300391?dgcid=rss_sd_all).)

[nemonominem]

Nick, you may be missing part of the logic. The probability of a cornovarius leak with contamination chain to the Wuhan games (and nowhere else) is much smaller than the probability of a coronavirus leak from a US lab.

I started with 1/4 of 1 leak event per 10 years from a US lab (one leak every 40 years for one BSL3 lab). The very odd directed chain to the Wuhan Games reduces that probability by roughly 1/20 (generously), based on all the requirements ((1) some asymptomatic personnel linked to a US lab to go to the Wuhan games, (2) asymptomatic petering out anywhere else). The unknown chain to the Wuhan outbreak (which the Chinese would have likely found out by now) reduces it by a further 1/2. Then you get the 1/160 of 1 event in 10 years.

[nemonominem]

On the number of biolabs in China, please read the very good article by Yuan Zhuming (Journal of Biosafety and Biosecurity). He gives information that is very difficult to get anywhere else:

He states that there were at least 42 BSL3s and 1,000 BSL2 in 2013. Plus some mobile labs.

[Update: this 2016 paper mentions 63 accredited BSL3/ABSL 3 labs in China. Based on this 80 P3 labs today is quite likely: https://www.centerforhealthsecurity.org/our-work/pubs_archive/pubs-pdfs/2016/National%20Biosafety%20Systems.pdf].

Coronaviruses are often handled at B2 level in China (no need at all to focus on a P4). I have not seen a new Chinese order to limit these to P3 or P4. Even if only a fraction of these 1,000 P2 handle coronaviruses this is still very risky (especially given what Yuan Zhiming says in his paper about the biosafety issues in these Chinese labs).

Same thing for the P3 labs - of these 42 how many are handling coronaviruses? By the way as 42 is the 2013 number, not sure why he can't give a more recent number. He also gives the number of BSL4 as 11 as of 2013 (53 - 42), far from the numbers reported elsewhere. I am not sure what to make of that.

Extract: In 2004, China’s central government launched a national BSL program that aimed at building a national BSL network comprising high-level BSLs as the nodes. High-level BSLs were then constructed and successfully operated.2 As of December 31st 2013, 53 BSLs, including 42 BSL-3s, had been fully accredited in China and more laboratories have completed the accreditation in recent years. In addition, more than 1000 BSL-2 labs are currently being operated in universities, research institutions, hospitals and R&D entrepreneurship centers.[2], [9], [12]

The number of these labs is going to keep growing fast. Each province is now required to have a P3 lab (with 30 more on the way just for Guangdong Province) and each major city a P2 lab. https://www.globaltimes.cn/content/1188916.shtml https://www.globaltimes.cn/content/1189394.shtml

By the way you can note the contradiction between the Chinese numbers. Yuan Zhiming (a credible CCP member) reports 42 P3 labs as of 2013. The 2016 paper mentioned above mention 63 P3 labs in 2016. Global Times (a party newspaper) reports only 20 P3 labs as of 2020!

[Nickleaton]

Any evidence that Corona virus are more prone than any other virus when it comes to a leak? No they all have the same probability of a leak.

Now I'm arguing that the number for the probability in the USA didn't match the reality. 1 in 140 years is too long a set of odds

On Mon, 6 Jul 2020 at 21:53, Gilles Demaneuf notifications@github.com wrote:

Nick, you need to look at the leaks of coronaviruses specifically. If there 10 P4 labs in the US but none is studying coronaviruses it does add to any probability of a coronavirus leak from these labs.

Also you should look at P2 to P4 labs in China. Most coronaviruses were studied at low P2 in Wuhan. SARS is a coronavirus and there were 4 leaks recently in China - that gives you a good idea of the calibration for Chinese labs studying coronaviruses. Also the article by Yuan Zhiming confirms that there risk in Chinese labs is very much present ( https://www.sciencedirect.com/science/article/pii/S2588933819300391?dgcid=rss_sd_all ).)

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-654459571, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAF7NEGH46Y2UA674IPHD4TR2I2U3ANCNFSM4MYUF6PQ .

-- Nick

[Nickleaton]

The probability of a cornovarius leak with contamination chain to the Wuhan games (and nowhere else) is much smaller than the probability of a coronavirus leak from a US lab.

Probably correct. Based on the papers that are linked on this site. However, given that they are lots of leaks in the US, the odds of 1 in 140 years is wrong.

On Mon, 6 Jul 2020 at 21:58, Gilles Demaneuf notifications@github.com wrote:

Nick, you are also missing part of the logic. The probability of a cornovarius leak with contamination chain to the Wuhan games (and nowhere else) is much smaller than the probability of a coronavirus leak from a US lab. The requirement for 'propagation chain to the Wuhan games and nowhere else' knocks that probability back by one of not 2 orders of magnitude. Wuhan is one of many places on earth. Why to Wuhan and nowhere else? You have to suppose military personnel in contact with asymptomatic cases going to Wuhan and that all the US cases are asymptomatic and peter out without a trace.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-654461615, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAF7NECIGXEMNNWBCI6GP6DR2I3GLANCNFSM4MYUF6PQ .

-- Nick

[nemonominem]

This 1/160 of 1 event in 10 year is based on the Chinese story of a leak from Fort Derek. Actually that's one lab in the US and only one, so we should look at its leak probability only. The source-outbreak probability used there was 1/4 of 1 leak in 10 year at BSL3. And yes the Chinese story is about coronaviruses, so we need to look at coronavirus leaks. I am not even sure that Fort Derek was studying coronaviruses, so that 1/4 of 1 in 10 year may be generous. Could well be zero!

[Nickleaton]

That was my point. The probability of a leak even at a well run lab is higher than 1:140 years. In a badly run lab it's even higher. That leads into the next step. What's the consequences of a leak. Could be very small, for example, the small pox leak in Birmingham in the UK. One person died. Or very serious like COVID 19

On Mon, 6 Jul 2020 at 22:59, Gilles Demaneuf notifications@github.com wrote:

This 1/160 is based on the Chinese story of a leak from Fort Derek. Actually that's one lab in the US and only one, so we should look at its leak probability only. One leak in 10 year for that lab is realistic.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-654486827, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAF7NEF44U5VR5ATIPKOXATR2JCNFANCNFSM4MYUF6PQ .

-- Nick

[nemonominem]

What are you trying to test Nick? Which question are you trying to answer?

I am explaining the rough calculation for the Chinese Fort Derek story where I clearly laid out the hypothesis. This was in the specific context of coronavirus leaks, which is made worse by GOF (mostly practiced in China actually) as this selects for rather nefarious viruses.

Mathematically it is difficult to fit an extreme event distribution with limited data. However if you consider that coronaviruses have the potential to cross species to humans, especially if selected through GOF for that very purpose, then a good assumption is that coronaviruses in GOF labs are not safe and are a prime suspect for this kind of outbreak. I also suspect that they were doing GOF at Bio level 3 at most, which would only make it worse.

I would also argue that by collecting 1,000s coronaviruses from caves you eventually run the risk of coming across one that is all too ready to jump species, without even GOF. And that's often handled typically at bio level 2 in the lab and rather up-hazardly during collection.

For US data, please see https://thebulletin.org/2019/02/human-error-in-high-biocontainment-labs-a-likely-pandemic-threat/. For a good probabilistic analysis see: https://thebulletin.org/2020/06/the-pandemic-risk-of-an-accidental-lab-leak-of-enhanced-flu-virus-unacceptably-high/ and https://armscontrolcenter.org/wp-content/uploads/2020/03/Quantifying-the-risk-9-17-Supplementary-material-at-end.pdf

[angoffinet]

I am not able to handle bayesian probabilities, and find arguments by Gilles quite compelling. To add a simple common sense comment, a leak of a highly contagious agent like SARS2 is more likely to have consequences than a leak of a less contagious agent (e.g. anthrax). Even viruses have different degree of contagiosity and airborne ones are much higher than enteric ones. What we have learned from SARS2 is that it is very contagious, probably much more than we thought at the beginning. How to factor this in the bayesian argument, I have no idea. Have a good day ! Andre Andre GOFFINET, MD, PhD Prof. em. Institute of Neuroscience University of Louvain, Belgium +32 (0) 473 899818

Le mar. 7 juil. 2020 à 02:49, Gilles Demaneuf notifications@github.com a écrit :

What are you trying to test Nick? I am explaining the rough calculation for the Chinese Fort Derek story where I clearly laid out the hypothesis. Which question are you trying to answer?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-654534864, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEX3H6LZBRLXVRAPZDCQGRTR2JWH3ANCNFSM4MYUF6PQ .

[Nickleaton]

The 1 in 140 years is clearly wrong. There have been leaks in the US at a higher rate. So either the model is wrong or the assumptions are wrong.

Perhaps there's also a pro US bias involved.

On Tue, 7 Jul 2020 at 01:49, Gilles Demaneuf notifications@github.com wrote:

What are you trying to test Nick? I am explaining the rough calculation for the Chinese Fort Derek story where I clearly laid out the hypothesis. Which question are you trying to answer?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-654534864, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAF7NEGOGI4DVMO75MTZ2E3R2JWH3ANCNFSM4MYUF6PQ .

-- Nick

[Nickleaton]

https://www.selectagents.gov/resources/FSAP_Annual_Report_2018_508.pdf

In 2018, the Federal Select Agent Program “received eight reports https://anthraxvaccine.blogspot.com/2020/04/select-agents-like-sars-cov-2-are.html of loss and 193 reports of a release of a biological select agent or toxin https://www.selectagents.gov/resources/FSAP_Annual_Report_2018_508.pdf.”

So the 1 in 140 years estimate for the US, doesn't stack up against the evidence.

Releases are frequent. Containment of something you can't see or detect [like radiation] is in my view next to impossible

On Tue, 7 Jul 2020 at 01:49, Gilles Demaneuf notifications@github.com wrote:

What are you trying to test Nick? I am explaining the rough calculation for the Chinese Fort Derek story where I clearly laid out the hypothesis. Which question are you trying to answer?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-654534864, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAF7NEGOGI4DVMO75MTZ2E3R2JWH3ANCNFSM4MYUF6PQ .

-- Nick

[nemonominem]

Sorry Nick, where do you get the 1/140 from? It could be a very conservative estimate. For instance this paper comes up with a quite similar conservative estimate for a single lab event per year, focussing on Avian flue GOF constructs and BSL 3 in the US (who knows what it is elsewhere): https://armscontrolcenter.org/wp-content/uploads/2020/03/Quantifying-the-risk-9-17-Supplementary-material-at-end.pdf

[angoffinet]

Gilles: I like that paper, even though I am not able to understand the statistical inferences techniques fully. Do you allow me to post the link on twitter and can I refer to you as source?

Andre GOFFINET, MD, PhD Prof. em. Institute of Neuroscience University of Louvain, Belgium +32 (0) 473 899818

Le mar. 7 juil. 2020 à 10:54, Gilles Demaneuf notifications@github.com a écrit :

Sorry Nick, where do you get the 1/140 from? It could be a very conservative estimate. For instance this paper comes up with a quite similar conservative estimate for a single lab event per year, focussing on Avian flue GOF constructs and BSL 3 in the US (who knows what it is elsewhere): https://armscontrolcenter.org/wp-content/uploads/2020/03/Quantifying-the-risk-9-17-Supplementary-material-at-end.pdf

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-654703123, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEX3H6MTDXFWFURODO55KELR2LPFBANCNFSM4MYUF6PQ .

[nemonominem]

Yes Andre, feel free to share it - I don't think that the author (Lynn C. Klotz, Center for Arms Control and Non-proliferation) would mind.

[nemonominem]

By the way, project-evidence.github.io is starting to get trolled. Their tactic seem to be to create many entries (see all the issues created by linomay https://github.com/Project-Evidence/project-evidence.github.io/issues) or to spurt out the usual Chinese propaganda in just slightly subtler ways (see https://github.com/Project-Evidence/project-evidence.github.io/issues/176).

[angoffinet]

Thank you. It is also my impression, sadly enough. The site has become too ramified and I cannot use it well (I do not know github architecture). Also, some comments are clearly biased to promote chinese official views. That's why I now prefer to use twitter where lots of interesting discussions are presented. Take care Andre

Andre GOFFINET, MD, PhD Prof. em. Institute of Neuroscience University of Louvain, Belgium +32 (0) 473 899818

Le mer. 8 juil. 2020 à 22:40, Gilles Demaneuf notifications@github.com a écrit :

By the way, project-evidence.github.io is starting to get trolled. Their tactic seem to be to create many entries (see all the issues created by linomay https://github.com/Project-Evidence/project-evidence.github.io/issues) or to spurt out the usual Chinese propaganda in just slightly subtler ways (see #176 https://github.com/Project-Evidence/project-evidence.github.io/issues/176 ).

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Project-Evidence/project-evidence.github.io/issues/31#issuecomment-655745031, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEX3H6NM3RF4XCLDH7EKGXDR2TKVHANCNFSM4MYUF6PQ .

[research-project-cov]

Peter Daszak now says RatG13 is a "consensus" virus... what do you understand? (50-55mn) https://youtu.be/Et3CHcteWNw