Genomic wastewater surveillance of human and animal influenza A viruses in California during the 2024-2025 flu season
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background
Wastewater genomic surveillance provides an opportunity to detect human and animal influenza A virus (IAV). We aimed to implement an IAV genomic surveillance framework agnostic to subtype, which enables recovery of IAV from multiple hosts and estimation of proportions across subtypes.
Methods
We conducted IAV genomic surveillance in wastewater during the 2024-2025 flu season at multiple sites in California and compared these data with available human clinical IAV sequences and test positivity. We applied a custom whole-genome, multi-host IAV probe enrichment panel and adapted our custom expectation-maximization (EM) algorithm to deconvolute IAV mixtures in wastewater and infer subtype relative abundances. Absolute IAV concentrations were quantified using RT-PCR-based assays. H5N1 wastewater and clinical sequences were further characterized by constructing a whole-genome maximum-likelihood phylogenetic tree. Finally, we performed variant analysis to examine amino acid substitutions detected in wastewater.
Findings
Our IAV probe enrichment method and EM algorithm successfully enriched all eight segments of three circulating IAV subtypes and accurately estimated subclade relative abundances for mixed IAV samples. Seasonal human H1N1pdm09 and H3N2 were detected throughout the study period from both wastewater and clinical sequencing data, with H1N1 subclades 6B.1A.5a.2a.1 and 6B.1A.5a.2a co-circulating, and H3N2 dominated by subclade 3C.2a1b.2a.2a.3a.1. Wastewater surveillance consistently detected H5N1 clade 2.3.4.4b across three monitored wastewater sites, while clinical H5N1 detections, from anywhere in CA, were sporadic and rare. Whole-genome phylogenetic analysis revealed that wastewater H5N1 sequences clustered with reference sequences associated with dairy cow and avian infections, while all human clinical H5N1 sequences clustered exclusively with reference sequences associated with dairy cow infections. Amino acid substitutions were identified across viral segments, and no mutations associated with mammalian adaptation were observed from wastewater samples.
Interpretation
When IAV concentrations were dominated by seasonal human subtypes rather than H5N1, subtype patterns aligned between wastewater and clinical data. While sequencing IAV in wastewater was unable to distinguish if H5N1 detections were due to human or animal infections, it was able to provide clade-level information about H5N1 found in wastewater that could be useful in the future. Wastewater genomic surveillance can complement clinical surveillance, increasing ability to detect all circulating IAV subtypes and enhancing public health preparedness from a One Health perspective.
Funding
UCOP Lab Fees CRT Award (L22CR4507) and NIH R00 Award (4R00GM144747-03)
Research in context
Evidence before this study
Sequencing IAV in wastewater has been shown to detect circulating subtypes from human and animal hosts, with potential to improve public health response, enhance surveillance for novel pandemic threats, and inform vaccine design. However, methodological improvements are needed for this potential to be fully realized. Targeted enrichment methods designed for whole-genome sequencing from wastewater can substantially improve genome coverage and sensitivity for low-abundance IAV. Approaches have included tiled amplicon, universal amplicon, and probe-capture enrichment. Tiled-amplicon methods provide high sensitivity and specificity but are less tolerant to sequence mismatches, and typically restrict primer design to a limited set of segments and a narrow range of subtypes. The universal amplicon approach was designed for clinical sequencing, where intact genomes from single isolates are available, and previous studies showed low recovery when applied to wastewater samples. In contrast, probe-capture enrichment tolerates mismatches and is therefore more resilient to RNA degradation and better suited to capture novel or divergent variants. However, current commercial probe-capture panels are pan-viral, which reduces sensitivity for IAV due to the high number of reads corresponding to other higher-abundance targets. The segmented nature of the IAV genome introduces additional challenges for downstream bioinformatic analysis. Existing deconvolution tools for wastewater sequencing have been developed primarily for SARS-CoV-2 and are not readily applicable to IAV. When applied to IAV, existing tools often demix only one subtype at a time, inferring clade-level relative abundance for that specific subtype. This limitation arises because different IAV subtypes require distinct reference genomes, unlike SARS-CoV-2, where a single reference can support alignment of multiple variants. Given that wastewater samples often contain multiple IAV subtypes simultaneously, it would be more useful to deconvolute clade-level relative abundances for all subtypes in parallel, underscoring the need for a more robust method tailored to the genomic complexity of IAV in wastewater.
Added value of this study
To our knowledge, this is the first study to apply an IAV-specific probe-capture enrichment panel to monitor IAV in wastewater, targeting all eight segments of the genome and subtypes across human, avian, dairy cattle, and other mammalian hosts. Unlike studies that prioritize either high coverage of very few subtypes or broad detection with limited genomic resolution, our study achieves a balance between capturing multiple IAV clades and maintaining reasonable coverage across segments. Additionally, we adapted a probabilistic expectation-maximization (EM) model to infer clade-level relative abundances from wastewater sequencing data. Our method includes imputation and retains fine-scale genomic differences, reducing bias when genome coverage is incomplete, which is a common scenario for low-abundance IAV in wastewater. While several bioinformatic tools have been developed to deconvolute SARS-CoV-2 mixtures in wastewater, comparable approaches have not been tailored for IAV, whose segmented genome and extensive subtype diversity pose unique challenges. Our framework fills this gap by enabling robust subclade-resolution surveillance across multiple subtypes and hosts using sequence data from wastewater samples.
Implications of all the available evidence
Our findings demonstrate a framework for wastewater genomic surveillance for tracking human IAV while also detecting animal-associated IAV. From a public health perspective, multi-host wastewater genomic surveillance offers two key benefits. First, it strengthens existing human IAV surveillance, because sequencing of clinical flu samples is limited and may not be representative of circulating subtypes, especially across regions with differing resources. Wastewater surveillance captures a larger and less biased sample population and has the potential to detect and track new subtypes more efficiently than clinical surveillance. Second, the ability to detect animal-associated IAV in wastewater supports pandemic preparedness by enabling monitoring for IAV of zoonotic potential and of genetic markers indicative of potentially increased human adaptation. This is especially beneficial where active surveillance of animal hosts is limited.
