UTILIZING NEXT GENERATION SEQUENCING AND MODELING APPROACHES TO ADVANCE MICROBIAL FOOD SAFETY

Abstract

The globalization of food production and trade has heightened the risk of foodborne outbreaks, emphasizing the urgent need for rapid and precise pathogen detection methods. Advances in whole genome sequencing (WGS) and metagenomic sequencing have transformed food safety surveillance, offering unparalleled resolution for pathogen identification, source attribution, and microbial community profiling. Yet, challenges remain in applying these technologies to complex food matrices and environmental samples. The overall goal of this research was to address these gaps by integrating next-generation sequencing with novel bioinformatic and modeling approaches to advance precision metagenomics and comparative genomics. Targeted amplicon-based microbiome characterization combined with structural equation modeling to elucidate the ecological and environmental factors that influence Listeria prevalence within food processing environments. This approach linked microbial community structure and sanitation practices to pathogen persistence risk, providing an evidence-based framework for intervention. To improve specificity in pathogen attribution, evaluating DNA sketching–based genome indexing on whole genome sequences as a complementary alternative to traditional serotyping. Salmonella Muenchen, a polyphyletic serovar of public health significance, was resolved more accurately using the genome indexing tool bettercallsal, which enhances serovar discrimination through proximity-based clustering while preserving historical antigenic nomenclature. This framework outperformed antigen- and cgMLST-based in-silico tools. Building on this, long-read Oxford Nanopore sequencing was combined with genome indexing for near real-time serovar detection in quasi-metagenomic samples. bettercallsal accurately identified multiple Salmonella serovars with as few as ~1,000 reads, drastically reducing the time and sequencing burden compared to assembly-dependent approaches such as SeqSero2. This rapid, non-assembly workflow enables informed microbiological decision-making during enrichment. Collectively, these studies demonstrate the power of combining WGS, metagenomics, and modeling to enhance microbial surveillance, enabling earlier detection, improved source tracking, and stronger outbreak prevention strategies within the One Health framework.