he news
media are fawning over a new
report
in Cell that claims to be the “strongest evidence yet”
that Covid came from a wet market.
The authors used metagenomic sequencing of samples collected from various parts of the Wuhan wet market and ran a DNA evolutionary analysis of PCR fragments to try to determine which strain of the virus SARS-CoV-2 might have been present. They claim that the earliest strains, called the "A" and "B" lineages, were on the samples found in the market.
In one of those amazing coincidences that always seems to happen, Matt Ridley, author of Viral: The Search for the Origin Of Covid-19, debunked this exact same argument last week, saying that finding traces of virus in the market is meaningless.
It's starting to look like the global warming argument: no matter how many times their predictions are falsified by evidence, advocates just come out with new ones that predict something even worse. So science gets a little more politicized and we science bloggers have to take more time out from more interesting topics, like the science of unclogging a bathtub, to explain it. Once a topic becomes politicized, bad science drives out the good.
The task of a peer reviewer is fiendishly hard: you can't just accuse the author of inventing his results. Nor is it a valid criticism to point out that the authors have already published several papers making the same claim. What you look for is (1) inconsistencies in the results, (2) conclusions that don't follow, and (3) experimental designs that are not capable of proving their case.
This Cell paper is an easy one. The authors do a very nice job of measuring the virus in different locations. But the paper fails on both (2) and (3).
1. There is no way to know if the animals were actually infected. The authors say “the timing of when viral or host genetic material were shed in the market environment cannot be directly estimated.” So we don't know if they were in the same place at the same time as the virus which, in case anyone doesn't know, is needed for infection. If the animals were not really infected, a market ‘origin’ is excluded. So their real conclusion is this:
The publicly available genomic and epidemiological data from the start of the COVID-19 pandemic remain incomplete, and future data from this time could shed further light on hypotheses regarding its emergence.
We're up to our earballs in hypotheses. What we need is a study that can discriminate among them.
2. The findings are not new. They suggest that maybe raccoon dogs, masked palm civets, hoary bamboo rats, and Malayan porcupines could have been involved. The evidence is purely circumstantial: DNA from the virus and from those species was found in animal carts, a hair-feather removal machine, and in drains. That means there could have been virus in the market, which we already know.
3. A ‘market origin’ is impossible. There is zero chance that any virus ‘originated’ in a market. Viruses do not originate in markets. They evolve out there among wild animals. To prove the wet-market hypothesis, the authors would have to prove that the animals were infected with the virus in the wild and then brought to the market. Finding traces of a virus in the market proves only that it was there. It says nothing about how it got there.
4. The results are consistent with both hypotheses. The authors say their results are “consistent with a market origin.” These weasel words are intended to imply that the results are inconsistent with a ‘lab leak’ origin. But this is not true. Almost no one doubts that Covid, or some form of it, evolved among animals. Virus samples from the animals were brought into the lab and studied. As the published record shows, introducing furin cleavage sites to increase the pathogenicity and tropism of infectious viruses was a common practice in several labs in that country, not just the WIV. The Cell authors had no samples from any lab, and by now the PRC government has certainly cleaned the labs thoroughly enough that no such samples will ever be found.
Thus, the paper fails on criterion 3 because the study design is unable in principle to answer the question. It fails on criterion 2 because even the weak conclusion that their findings are “consistent” with a market origin does not follow from their results. More importantly, their data say nothing about a lab leak. They use typical ambiguous scientific language of uncertainty to cover themselves, but the news media always overlook this.
What would they have needed to prove their case? The dog that didn't bark is the lab samples. If lab samples taken at various times showed a clear time progression from the market to the lab, or vice-versa, it might help. But to actually prove the case, they'd need to find a population of animals in the wild where the original virus—complete with the observed humanized adaptations and furin cleavage site—existed endemically before it got to the market. Unfortunately, it's far too late now. If we found such a population of animals now, it would be impossible to know how or when they got the virus. All we can do now is speculate.
It may be understandable that biologists would be almost desperate to convince us that science was not responsible for so many deaths. It has certainly damaged the reputation of biology, of the NIH, and of science and public health in general. As a result of the disastrous way they handled Covid, their efforts to fight the next pandemic will be much more difficult. But covering up their failures won't rebuild the lost trust.
Does it matter whether COVID came from a lab? Only in the sense that what matters is that we know the truth. We didn't know it before, and we don't know it now.
sep 20 2024, 7:52 am
