Open AccessEfficient and effective single-step screening of individual samples for SARS-CoV-2 RNA using multi-dimensional pooling and Bayesian inference
Author(s) -
J Sobczyk,
Michael T. Pyne,
Adam P. Barker,
Jeanmarie Mayer,
Kimberly E. Hanson,
Matthew H. Samore,
Rodrigo Noriega
Publication year - 2021
Publication title -
journal of the royal society interface
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.655
H-Index - 139
eISSN - 1742-5689
pISSN - 1742-5662
DOI - 10.1098/rsif.2021.0155
Subject(s) - pooling , protocol (science) , computer science , bayesian probability , inference , bayesian inference , sample (material) , computational biology , data mining , machine learning , medicine , artificial intelligence , biology , pathology , chemistry , alternative medicine , chromatography
Rapid and widespread implementation of infectious disease surveillance is a critical component in the response to novel health threats. Molecular assays are the preferred method to detect a broad range of viral pathogens with high sensitivity and specificity. The implementation of molecular assay testing in a rapidly evolving public health emergency, such as the ongoing COVID-19 pandemic, can be hindered by resource availability or technical constraints. We present a screening strategy that is easily scaled up to support a sustained large volume of testing over long periods of time. This non-adaptive pooled-sample screening protocol employs Bayesian inference to yield a reportable outcome for each individual sample in a single testing step (no confirmation of positive results required). The proposed method is validated using clinical specimens tested using a real-time reverse transcription polymerase chain reaction test for SARS-CoV-2. This screening protocol has substantial advantages for its implementation, including higher sample throughput, faster time to results, no need to retrieve previously screened samples from storage to undergo retesting, and excellent performance of the algorithm's sensitivity and specificity compared with the individual test's metrics.