Search bioRxiv⌕ Search

Biology subjects

Lison, A.

Publications and source records attributed to Lison, A..

2 recordsLinked to original sources

Post-sampling degradation of viral RNA in wastewater impacts the quality of PCR-based concentration estimates

Successful wastewater-based infectious disease surveillance programs depend on regular, reliable molecular detection of nucleic acids in municipal wastewater systems. This process is challenged by the gradual degradation of the viral content of the wastewater over time. Testing protocols are complex and often cannot be performed on site, resulting in delays between collection and testing. The evidence of continued degradation of viral RNA when stored at low temperatures is currently limited to a small number of studies with mixed results. Taking advantage of variable delays between sample collection and processing, we used a Bayesian temporal model and data from two winter periods of a national surveillance program in Switzerland to determine the rate at which the measured viral concentrations of four respiratory viruses declined as a result of RNA degradation between sample collection and processing. We found evidence of substantial degradation between the collection and processing of samples with daily rates of up to -0.28 (-0.38 - -0.19 95% CrI). We established that reduction in viral concentrations resulting from post-sampling degradation was responsible for a number of measurements falling below quantifiable limits. For one treatment plant, we estimate that 39 measurements fell below the limit of detection due to RNA degradation over the course of a single season. Measurements are more likely to be lost early in the seasonal epidemic when concentrations are still low. This delays consistent reliable measurement and sets back epidemiological assessments relevant to public health management strategies.

microbiology↗

Improving inference in wastewater-based epidemiology by modelling the statistical features of digital PCR

Digital polymerase chain reaction (dPCR) is a powerful technique for quantifying gene targets in environmental samples, with various applications such as biodiversity monitoring and wastewater-based epidemiology. However, statistical analyses of environmental dPCR data often assume, explicitly or implicitly, that concentration measurements have a normal or log-normal error structure, which does not reflect the underlying partitioning statistics of dPCR. Using simulations and real-world environmental data, we show that (log-)normality assumptions are violated for dPCR measurements, leading to inaccurate estimates of gene concentrations and underlying biological processes. To enable reliable analyses of environmental dPCR data, we present a dPCR-specific likelihood model that accounts for concentration-dependent measurement noise and non-detects as characteristic of dPCR assays. We demonstrate that this approach overcomes biases in inference from environmental data, such as estimating free-eDNA decay in seawater or pathogen transmission from wastewater monitoring. Our method is implemented in the R packages "dPCRfit" for regression analyses and "EpiSewer" for wastewater surveillance.

bioinformatics↗