summary - Argues SARS-CoV-2’s genome - a bat-coronavirus (RaTG13-like) backbone combined with a receptor-binding domain nearly identical to a pangolin coronavirus, plus a furin cleavage site “previously unseen in other SARS-like CoVs” - is more consistent with laboratory assembly than natural recombination. Central claim: the 12-nucleotide furin-site insertion “is not in frame” with the pattern expected from natural polymerase slippage/recombination, which the authors read as more consistent with deliberate site-directed mutagenesis (e.g. experiments testing pangolins as an intermediate host, or gain-of-function-style furin-site insertion work of the kind later described in the leaked DEFUSE proposal). Directly contests Andersen et al. 2020’s natural-evolution reading of the same two genomic features.
relevance_note - The most-cited peer-reviewed argument that the furin site looks engineered - the direct published counter to Andersen et al. 2020 (separate source node, this slice) on the same evidence: same underlying SARS-CoV-2/RaTG13/pangolin-CoV sequence data, opposite reading of the furin-site insertion mechanism. Flagged per the recurring-dataset note - not an independent dataset from Andersen 2020’s.