[inciampati]

The slow mutation rate helps us date the "spillover". But it isn't relevant to my argument. If we saw one, two, and three mutations per strain in late December, it means their common ancestor couldn't have been more than a few months prior. If we saw tens, or different clades with specific geographic distribution patterns, then we might look at those to understand where the introduction to humans was. My point is that this introduction probably happened in Wuhan itself, or there would have had to be a complex and highly improbable series of transmissions from a host animal (presumably a bat 1000km away from Wuhan) which spawned no daughter strains. And thanks to the high fidelity replication process of the virus, these strains would be exactly the same in phenotype as the one that showed up in Wuhan. Why did we miss them?

SARS-CoV-1 underwent clear adaptation in humans that was observed through sequencing. In observation, the mutation pattern over time wasn't initially driven by a stable molecular clock. There were easily-discoverable mutations which yielded phenotypic gain. The virus is high fidelity, but there is enough of it that it's exploring a wide range of possible options at every infection. For a virus close to optimum, very few of these will yield a benefit, and so we see mutations that are mostly synonymous that occur at a clock like rate. Sound familiar?