Importance of quantifying the number of viral reads in metagenomic sequencing of environmental samples from the Huanan Seafood Market

Abstract

Abstract In March 2023, the Chinese CDC publicly released raw metagenomic sequencing data for environmental samples collected in early 2020 from the Huanan Seafood Market. Prior to that data release, some scientists had suggested that these samples could be informative for establishing if animals such as raccoon dogs had been infected with severe acute respiratory syndrome virus 2 (SARS-CoV-2). However, no one had analyzed how much SARS-CoV-2 was actually present in the metagenomic sequencing data. After the raw data became available, I fully analyzed the abundance of both viral and animal genetic material in the samples. That analysis, which was published in Virus Evolution, found that the SARS-CoV-2 content of most samples was very low and that the abundance of SARS-CoV-2 was most strongly associated with animals such as largemouth bass that are not plausible candidates for having been infected. Based on these results, I concluded that the metagenomic content of the samples was not informative for determining if any non-human animals in the market had been infected with SARS-CoV-2. One of the authors of an earlier study of these samples, Florence Débarre, recently submitted a response to my paper. Here, I reply in turn to explain why it is important to quantify the abundance of viral material before drawing conclusions from metagenomic sequencing. I also report new analyses of other animal coronaviruses in the samples and show that material from some other animal coronaviruses is much more abundant than SARS-CoV-2 in samples collected on the date when most wildlife stall sampling was performed. I further show that material from some of these animal coronaviruses is associated with the animals they probably infect but that no such association exists for SARS-CoV-2. Overall, these new analyses further emphasize the importance of quantifying the actual amount of viral material in metagenomic samples and underscore why the environmental samples from the Huanan Seafood Market are not informative for determining if any non-human animals were infected with SARS-CoV-2.