Untargeted sequencing of a city's sewage can detect a range of viruses (even novel ones), providing early warning for pandemics. Its value depends on its sensitivity (P2RA): how strong is a virus's sewage signal at a given prevalence level? During my Frontier Biodefense Fellowship at Pivotal Research, mentored by Lennart Justen (MIT, Broad Institute), I refined SecureBio's P2RA estimates for respiratory viruses using more nasal swab data and an updated Bayesian model that lets viral prevalence vary week by week. For the seven most observed viruses, the range of plausible P2RA values tightened by a factor of 2–5. I will continue working on the project during a two-month extension.
Here is the poster (click it for the PDF version):
