Back

Introducing NOMAD: A Field-Deployable Workflow for Rapid, On-Site Microbiome Analysis of Raw Bovine Milk Using Full-Length 16S rRNA Sequencing

Parreira, V. R.; Precious, M.; Taggar, G.; Singh, D.; Goodridge, L.

2025-09-30 genomics
10.1101/2025.09.29.679195 bioRxiv
Show abstract

Raw bovine milk hosts a diverse microbiota that profoundly influences dairy product quality, safety, and shelf-life. However, current surveillance methods are time-intensive and often lack the taxonomic resolution needed for effective risk mitigation. To address this, we developed NOMAD (Nanopore-based On-site Microbiome Analysis of Diversity), a field-deployable workflow for rapid, high-resolution characterization of the raw milk microbiome using full-length 16S rRNA gene sequencing via Oxford Nanopore Technologies. Milk samples collected from a commercial dairy operation were processed using eight DNA extraction protocols, with Method (incorporation of EDTA and TE buffer) emerging as the optimal approach for microbial richness and DNA yield. Sequencing was performed on a MinION Mk1B platform, and bioinformatic analyses revealed that a 4-hour run was sufficient to recover >90% of total community richness, with stable alpha and beta diversity metrics by this timepoint. The complete workflow, including DNA extraction, library preparation, sequencing, and analysis, was completed in 10.5 hours, enabling same-day microbiome profiling in farm-adjacent settings. Comparative analysis showed strong agreement with established milk microbiome studies, while full-length reads enhanced resolution of spoilage-associated taxa such as Pseudomonas spp. and Streptococcus spp. The NOMAD platform offers a powerful and practical tool for near real-time microbiological surveillance in the dairy industry, supporting proactive quality control and improved food safety outcomes.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.