Inferring viral variants and fitness advantages from longitudinal wastewater sequencing data

FitNMF infers mutational signatures, relative abundances, and fitness advantages of circulating SARS-CoV-2 variants directly from wastewater sequencing, identifying emerging variants de novo across six Swiss plants from 2021–2024.

• Why it matters: Wastewater is a pooled, fragmentary mixture that blocks complete genome assembly, and existing pipelines struggle to detect variants absent from clinical sequencing.
• What they did: FitNMF combines viral dynamics with non-negative matrix factorization to exploit temporal structure and variant competition, validated on simulated data and Swiss wastewater time series from 2021–2024 across six treatment plants.
• The result: The model can identify and quantify emerging variants de novo, enabling real-time SARS-CoV-2 variant surveillance where clinical sequencing is limited.

AI summary of the abstract · may contain errors · not medical advice · read the paper.