Early Detection of Novel SARS-CoV-2 Variants from Urban and Rural Wastewater through Genome Sequencing and Machine Learning.

Genome sequencing from wastewater has emerged as an accurate and cost-effective tool for identifying SARS-CoV-2 variants. However, existing methods for analyzing wastewater sequencing data are not designed to detect novel variants that have not been characterized in humans. Here, we present an unsupervised learning approach that clusters co-varying and time-evolving mutation patterns leading to the identification of SARS-CoV-2 variants. To build our model, we sequenced 3,659 wastewater samples collected over a span of more than two years from urban and rural locations in Southern Nevada. We then developed a multivariate independent component analysis (ICA)-based pipeline to transform mutation frequencies into independent sources with co-varying and time-evolving patterns and compared variant predictions to >5,000 SARS-CoV-2 clinical genomes isolated from Nevadans. Using the source patterns as data-driven reference "barcodes", we demonstrated the model's accuracy by successfully detecting the Delta variant in late 2021, Omicron variants in 2022, and emerging recombinant XBB variants in 2023. Our approach revealed the spatial and temporal dynamics of variants in both urban and rural regions; achieved earlier detection of most variants compared to other computational tools; and uncovered unique co-varying mutation patterns not associated with any known variant. The multivariate nature of our pipeline boosts statistical power and can support accurate and early detection of SARS-CoV-2 variants. This feature offers a unique opportunity for novel variant and pathogen detection, even in the absence of clinical testing.

Identification of a rare SARS-CoV-2 XL hybrid variant in wastewater and the subsequent discovery of two infected individuals in Nevada.

The identification of novel SARS-CoV-2 variants can predict new patterns of COVID-19 community transmission and lead to the deployment of public health resources. However, increased access to at-home antigen tests and reduced free PCR tests have recently led to data gaps for the surveillance of evolving SARS-CoV-2 variants. To overcome such limitations, we asked whether wastewater surveillance could be leveraged to detect rare variants circulating in a community before local detection in human cases. Here, we performed whole genome sequencing (WGS) of SARS-CoV-2 from a wastewater treatment plant serving Las Vegas, Nevada in April 2022. Using metrics that exceeded 100× depth at a coverage of >90 % of the viral genome, we identified a variant profile similar to the XL recombinant lineage containing 26 mutations found in BA.1 and BA.2 and three private mutations. Prompted by the discovery of this rare lineage in wastewater, we analyzed clinical COVID-19 sequencing data from Southern Nevada and identified two cases infected with the XL lineage. Taken together, our data highlight how wastewater genome sequencing data can be used to discover rare SARS-CoV-2 lineages in a community and complement local public health surveillance.