Multi-country and intersectoral assessment of cluster congruence between different bioinformatics pipelines for genomics surveillance of foodborne bacterial pathogens

Abstract

AbstractFood and waterborne disease (FWD) surveillance requires Whole-Genome Sequencing (WGS)-based systems following a One Health approach. However, different laboratories employ different WGS pipelines in their routine surveillance activities, casting doubt on the comparability of their results and hindering optimal communication at intersectoral and international levels. Through a collaborative effort involving eleven European institutes across seven countries and spanning the food, animal and human health sectors, we aimed to assess the inter-laboratory comparability of WGS clustering results for four important foodborne pathogens:Listeria monocytogenes,Salmonella enterica,Escherichia coliandCampylobacter jejuni. Each participating institute (n=9) applied its surveillance pipeline over the same WGS datasets (>2000 isolates per species), and, for each pipeline, genetic clusters were identified at each possible allele/SNP distance threshold. Inter-pipeline clustering congruence was assessed by calculating a “Congruence Score” (relying on Adjusted Wallace and Adjusted Rand coefficients) across all resolution levels, followed by an in-depth comparative analysis of cluster composition at outbreak level. An additional cluster congruence assessment was performed between WGS and traditional typing, which, depending on the species, included Sequence Type (ST), Clonal Complex (CC) and/or serotype. Our results revealed a general high concordance between allele-based pipelines at all resolution levels for all species, except forC. jejuni, where the different resolution power of available allele-based schemas led to marked discrepancies. Still, this study identified non-negligible differences in allele-based pipeline performance for outbreak cluster detection, suggesting that a threshold flexibilization is important for the detection of similar outbreak signals by different laboratories. These results, together with the observation that different STs, CCs and serotypes exhibit remarkably different genetic diversity, should inform future threshold selections for outbreak case definitions. In conclusion, this study provides valuable insights into the comparability of pipelines commonly used for routine genomics surveillance, and reinforces the need, while demonstrating the feasibility, of conducting continuous and comprehensive WGS pipeline comparability assessments. Ultimately, it opens good perspectives for a smoother international and intersectoral cooperation and communication towards a sustainable and efficient One Health FWD surveillance.