Emily Somerset, PhD candidate, Faculty of Arts & Science

Supervised by Patrick Brown, Faculty of Arts & Science, Monica Alexander, Faculty of Arts & Science, and Lennon Li, Dalla Lana School of Public Health

Project Title: A spatio-temporal framework for estimation and interpretation of wastewater viral signals

Project Summary: Wastewater-based surveillance stands as a forefront method for disease monitoring, offering a real-time, representative depiction of disease presence, irrespective of access to clinical testing or symptom status. Currently, a wide range of pathogens, including SARS-CoV-2, influenza, hepatitis, and respiratory syncytial virus (RSV), are being monitored through wastewater, highlighting its potential for proactive public health management.

Public health researchers want reliable estimates of the viral signal and its direction, for public dissemination. Wastewater modeling presents challenges due to low signal-to-noise ratios, unequal sampling, temporal correlation, and signal variability across communities. Improved statistical methods are needed to separate signal from noise and articulate uncertainty in wastewater trends. Simultaneously, to ensure effective public dissemination, these advanced models need to run in a timely manner.

We propose a flexible, fast, Bayesian hierarchical modeling framework adaptable to all pathogens, sampling schedules and availability of control measurements for normalization. Moreover, we propose modelling viral signals as differentiable processes, enabling direct estimation of their derivatives. This way, viral signals can be conveniently presented alongside their derivatives, with uncertainty intervals, as instantaneous indicators of change. We will share our established models with the community through an open-source, user-friendly R package.