O-2 - No Huanan animal-origin sample (457 samples, 18 species) tested SARS-CoV-2-positive
No animal tissue itself tested positive, so this dataset alone cannot exhibit a directly infected animal. Sampling of animal material began 18 Jan 2020, after the 1 Jan closure and disinfection and after most live animals had been removed.
Link to original
Why this is evidence
H-1 is the only member that predicts an infected animal could have been present at the market at all, so it alone assigns non-trivial probability to a positive animal sample; H-5 and H-6 predict zero positives with near-certainty. The strength of the miss is heavily discounted by the sampling timeline (O-3), which is why that caveat is carried by an argument rather than by this edge.