Start with the denominator, because the published account does not give one. We learn which manuscripts came up positive, and that in several of them more than one page did [4], but not how many folios were screened to get there, and not what share of screened folios returned sheeppox reads [3]. That fraction is the whole question. At one positive in twenty ordinary codices, libraries become a sampling frame where every sample arrives with a date and a place of production attached, which excavated bone rarely offers. At one in two hundred, each genome is a costly prize and the archive stays a curiosity.
The result I would want tested next is the virus on calfskin and goatskin [5]. The authors offer two readings, rare cross-species infection or contamination during parchment making, and those are not interchangeable for anyone who wants to treat a positive page as an epidemiological data point. If sheeppox DNA travelled in the lime pits and scraping benches of a workshop, a positive folio records the workshop rather than the animal, and the neat pairing of a genome with one herd at one date comes apart. Several positive pages in a single book [4] fits an infected flock and also fits a shared bath of process water. There is a clean test: gatherings that mix species, or books with no sheep skin in them at all, should carry the signal if the craft rather than the animal is the source.
The dating carries its own caution. Placing the divergence of the sheeppox, goatpox and lumpy skin disease lineages between 11,500 and 3,700 years ago [6] is a window 7,800 years wide [1], and its young edge sits where the oldest direct evidence sits, the Bronze Age dental remains from steppe pastoralist settlements at more than 3,700 years [7]. Molecular clocks behave that way when one old tip is doing most of the work; more ancient sequences, not more modern ones, are what would narrow it. The manuscript material cannot supply that depth. The York Gospels at roughly a millennium [2] sit inside about the most recent quarter of the 3,700-year record [2]. Teeth give the depth [8]; books could give the density.
What the parchment does establish is that an animal pathogen's DNA survives the making of a book, which is the claim the lead author extends to archives and libraries generally [9]. Sheeppox is still worth the trouble: it moves fast through a flock, kills young animals at high rates, and cuts wool, meat and milk yields [10], and the study's authors note that it is still being fought [11]. For a disease like that, a dated series of genomes reaching back three and a half millennia is a better calibration than another modern isolate.
