3 ms·

Of course, it can be evolution of the virus like in all cases before SARS CoV2, but no animal reservoir, where this evolution happened, is found. Viruses are n

Of course, it can be evolution of the virus like in all cases before SARS CoV2, but no animal reservoir, where this evolution happened, is found. Viruses are not alive, so evolution cannot be performed without a chain (tree) of host animals and/or humans. However, in this case, it looks like it was jump, not evolution. At one point of time we see the virus in wild without ability to easily infect humans, then at another point in time we see the almost same virus, but perfectly adapted to infect humans, and nothing between them. If this was an evolution in the wild, then thousands, maybe millions, of "almost SARS CoV2" cases should surface after years of search. They are not found, thus evolution happened in a closed environment, like a secret cave or a biolab.

__MatrixMan__ 2y ago
> If this was an evolution in the wild, then thousands, maybe millions, of "almost SARS CoV2" cases should surface after years of search I question that assumption. Evolution happens in tiny bursts (founder organisms, population bottlenecks, etc) in between which are relatively long periods of Hardy-Weinberg equilibrium (i.e. periods where evolution is not happening). That's why when we look into the fossil record we're (typically) able classify our findings as one of several stable species, rather than some sort of smooth spectrum in which everything is equally represented a transitional state to something else. There's a very strong sampling bias where you're much more likely to find an organism in equilibrium and not an individual that was part of whatever transition you're interested in. If this were not the case, the existence of evolution would've been a no-brainer rather than a discovery that biologists had to fight to establish. In the last 4 years there have been 20 variants of concern, but currently there are only two (https://www.ecdc.europa.eu/en/covid-19/variants-concern https://www.ecdc.europa.eu/en/covid-19/variants-concern). So on average these populations only last something like six months before they're outcompeted by a new variant--and that's six months for populations in equilibrium. If you're looking to collect a sample in a non-equilibrium state, it seems likely to me that your window for doing so might only last a few weeks. So they're looking for the missing link, and they haven't found it yet. But it's not clear how likely they are to ever find it, so the jump from "they haven't found it" to "it was never in nature" is not warranted. One experiment that could be done, which I think would shed light on the situation, would be sample modern humans in an attempt to find the first variant (discovered in 2019). If we find it easily, maybe we should expect to find its ancestors relatively easily, supposing they're out there (never mind that animals are harder to work with than humans). Then again, it may be extinct, in which case we should probably not hold our breaths about finding its ancestor either. But sequencing experiments cost on the order of $100-$1000 to do, so neither my proposed experiment nor the quest to find the reservoir species are likely to happen at a scale of > thousands, maybe millions, of ... cases Sure, you might get more than one virus per sample, but typically the way to sort that out is to align the reads (typically 100-200 nucleotides long) to a reference genome. It works well when your sample came from one individual with genetically identical cells. I'm not sure how quickly things get murky when you've got a pile of reads from a sample with hundreds or thousands of genetically dissimilar viral particles, plus whatever non-viral DNA ended up in the sample, but I'd be surprised if you could get more than 10 or 20 reliable alignments out of it. Eventually your error bars are going to get too big and you're not going to know which data goes which which sites. I haven't worked with an assay like that before but I'm positive that there are going to be limits to how much you can squeeze out of a single run.
- oneshtein 2y ago
  Viruses are evolving at somewhat constant rate, because constant rate of mutations is required to bypass immune system. Viruses must mutate or they will be eliminated by immune system of their hosts. So, we can multiple the rate of mutations by number of hosts to get average speed of evolution. By looking at genome of virus strain and knowing rate of mutations and number of hosts, we can calculate the time passed between them. By testing immune response for variants of the virus, we can find the person, who was infected first, AKA «patient zero», then look for a nearby cave and find the animal, which started the infection. So, by prediction, the infection in humans started at Sep-Oct 2019. First known patient in China was infected in Nov 2019: too late for the «patient zero». It means that infection, most likely, originated outside of China, where original bat virus was found. So, we will NEVER find a cave in China with pre-covid19 virus, because infection started outside of China. 0 The cave with pre-covid virus is in China, but infection started outside of China. WTF? Explosion at «Vector» lab, Koltsevo, Novosibirsk, Siberia, Russia was at Sep 16 2019, so we have match here. +1 Joint training of Chinese from Wuhan and Russian military police was started at Oct 11, 2019. They cannot bring infection from China to Russia because first case in China was in November, but Chinese police can be infected in Russia and return to China with infection in late October-November. +1 First report in media about sudden increase in cases of atypical pneumonia (start of infection) in Krasnoyark Krai, Siberia (700 cases per week) was at Oct 16, 2019. +1 So, we have 0 points in favor of animal origin, and 3 points in favor of lab origin.
  - __MatrixMan__ 2y ago
    That really just looks like 0 points on either side to me. I find it quite plausible that it started outside of China, because why not, but if there was some kind of foul play involved I don't see why we should expect to be able to unravel the misinformation without having been personally involved. Basically none of the reporting on it was accurate. I had it two weeks prior to the first reported case in my state and was told that "everybody who needs a test can get one" and was unable to get a test because I wasn't sick enough to justify it (which is fine, but if that's how your surveillance works then your data is useless). It's a big cloud of lies and I just don't see the utility of peering into it and trying to assign blame when that same energy could be put towards being prepared for next time.

__MatrixMan__ 2y ago

> If this was an evolution in the wild, then thousands, maybe millions, of "almost SARS CoV2" cases should surface after years of search I question that assumption. Evolution happens in tiny bursts (founder organisms, population bottlenecks, etc) in between which are relatively long periods of Hardy-Weinberg equilibrium (i.e. periods where evolution is not happening). That's why when we look into the fossil record we're (typically) able classify our findings as one of several stable species, rather than some sort of smooth spectrum in which everything is equally represented a transitional state to something else. There's a very strong sampling bias where you're much more likely to find an organism in equilibrium and not an individual that was part of whatever transition you're interested in. If this were not the case, the existence of evolution would've been a no-brainer rather than a discovery that biologists had to fight to establish. In the last 4 years there have been 20 variants of concern, but currently there are only two (https://www.ecdc.europa.eu/en/covid-19/variants-concern https://www.ecdc.europa.eu/en/covid-19/variants-concern). So on average these populations only last something like six months before they're outcompeted by a new variant--and that's six months for populations in equilibrium. If you're looking to collect a sample in a non-equilibrium state, it seems likely to me that your window for doing so might only last a few weeks. So they're looking for the missing link, and they haven't found it yet. But it's not clear how likely they are to ever find it, so the jump from "they haven't found it" to "it was never in nature" is not warranted. One experiment that could be done, which I think would shed light on the situation, would be sample modern humans in an attempt to find the first variant (discovered in 2019). If we find it easily, maybe we should expect to find its ancestors relatively easily, supposing they're out there (never mind that animals are harder to work with than humans). Then again, it may be extinct, in which case we should probably not hold our breaths about finding its ancestor either. But sequencing experiments cost on the order of $100-$1000 to do, so neither my proposed experiment nor the quest to find the reservoir species are likely to happen at a scale of > thousands, maybe millions, of ... cases Sure, you might get more than one virus per sample, but typically the way to sort that out is to align the reads (typically 100-200 nucleotides long) to a reference genome. It works well when your sample came from one individual with genetically identical cells. I'm not sure how quickly things get murky when you've got a pile of reads from a sample with hundreds or thousands of genetically dissimilar viral particles, plus whatever non-viral DNA ended up in the sample, but I'd be surprised if you could get more than 10 or 20 reliable alignments out of it. Eventually your error bars are going to get too big and you're not going to know which data goes which which sites. I haven't worked with an assay like that before but I'm positive that there are going to be limits to how much you can squeeze out of a single run.

oneshtein 2y ago
Viruses are evolving at somewhat constant rate, because constant rate of mutations is required to bypass immune system. Viruses must mutate or they will be eliminated by immune system of their hosts. So, we can multiple the rate of mutations by number of hosts to get average speed of evolution. By looking at genome of virus strain and knowing rate of mutations and number of hosts, we can calculate the time passed between them. By testing immune response for variants of the virus, we can find the person, who was infected first, AKA «patient zero», then look for a nearby cave and find the animal, which started the infection. So, by prediction, the infection in humans started at Sep-Oct 2019. First known patient in China was infected in Nov 2019: too late for the «patient zero». It means that infection, most likely, originated outside of China, where original bat virus was found. So, we will NEVER find a cave in China with pre-covid19 virus, because infection started outside of China. 0 The cave with pre-covid virus is in China, but infection started outside of China. WTF? Explosion at «Vector» lab, Koltsevo, Novosibirsk, Siberia, Russia was at Sep 16 2019, so we have match here. +1 Joint training of Chinese from Wuhan and Russian military police was started at Oct 11, 2019. They cannot bring infection from China to Russia because first case in China was in November, but Chinese police can be infected in Russia and return to China with infection in late October-November. +1 First report in media about sudden increase in cases of atypical pneumonia (start of infection) in Krasnoyark Krai, Siberia (700 cases per week) was at Oct 16, 2019. +1 So, we have 0 points in favor of animal origin, and 3 points in favor of lab origin.
- __MatrixMan__ 2y ago
  That really just looks like 0 points on either side to me. I find it quite plausible that it started outside of China, because why not, but if there was some kind of foul play involved I don't see why we should expect to be able to unravel the misinformation without having been personally involved. Basically none of the reporting on it was accurate. I had it two weeks prior to the first reported case in my state and was told that "everybody who needs a test can get one" and was unable to get a test because I wasn't sick enough to justify it (which is fine, but if that's how your surveillance works then your data is useless). It's a big cloud of lies and I just don't see the utility of peering into it and trying to assign blame when that same energy could be put towards being prepared for next time.