I think you're referring to this paper (or other similar analyses): https://www.cell.com/cell/pdf/S0092-8674(24)00901-2.pdf https://www.cell.com/cell/pdf/S0092

I think you're referring to this paper (or other similar analyses): https://www.cell.com/cell/pdf/S0092-8674(24)00901-2.pdf https://www.cell.com/cell/pdf/S0092-8674(24)00901-2.pdf It found that viral RNA and DNA from potential zoonotic hosts (palm civets, raccoon dogs) occurred together in some samples from the Huanan Seafood Market. But there's no question that infected humans were present there, and nothing in their study can distinguish an infected palm civet from a healthy palm civet near an infected human. A different study by Jesse Bloom found that of all the animals present, viral RNA correlated most strongly with DNA from "catfish and largemouth bass": https://academic.oup.com/ve/article/9/2/vead050/7249794 https://academic.oup.com/ve/article/9/2/vead050/7249794 So were those the proximal hosts? Obviously not, since SARS-CoV-2 can't infect fish--the point is that this is all just noise. Live, unquestionably infected proximal hosts were found for both SARS-1 and MERS, within about a year of the first human cases. For SARS-CoV-2, we still have nothing beyond those tortured metagenomics.