T5-09 Whole Genome Sequencing-based Benchmarking of Subtyping Methods for Salmonella enterica Serotype Enteritidis

Tuesday, August 5, 2014: 11:00 AM
Room 111-112 (Indiana Convention Center)
Xiangyu Deng, University of Georgia, Griffin, GA
Nikki Shariat, The Pennsylvania State University, University Park, PA
Elizabeth Driebe, Translational Genomics Research Institute, Flagstaff, AZ
Beth Tolar, Centers for Disease Control and Prevention, Atlanta, GA
Eija Trees, Centers for Disease Control and Prevention, Atlanta, GA
Paul Keim, Translational Genomics Research Institute, Flagstaff, AZ
Wei Zhang, Illinois Institute of Technology, Bedford Park, IL
Edward Dudley, The Pennsylvania State University, University Park, PA
Patricia Fields, Centers for Disease Control and Prevention, Atlanta, GA
David Engelthaler, Translational Genomics Research Institute, Flagstaff, AZ
Introduction: Powered by whole genome sequencing (WGS) technologies, recent implementations of whole genome single nucleotide polymorphism typing (WGST) have led to significant improvements of both molecular subtyping and phylogenetic analyses, particularly for monophyletic bacterial pathogens. WGST allows discovery of SNPs across the entire bacterial genomes, therefore, provides superior subtyping resolution and phylogenetic accuracy, which make it ideal for benchmarking other subtyping methods.

Purpose: To compare the respective performance of commonly used and newly developed subtyping methods in delineating each individual outbreak under the guidance of the recently proposed phylogenetic framework and population structure of commonly circulating SE lineages in the United States. 

Methods: A cohort of 52 Salmonella enterica serotype Enteritidis (SE) isolates representing 16 major outbreaks and three sporadic cases in the United States between 2001 and 2011 were sequenced and subject to a retrospective investigation in combination with four subtyping methods. 

Results: Whole genome single nucleotide polymorphism typing (WGST) resolved all outbreak clusters and provided useful epidemiological information with robust phylogenetic inference. While both MLVA and CRISPR-MVLST yielded higher discriminatory power than PFGE, MLVA outperformed in delineating outbreak clusters whereas CRISPR-MVLST showed the potential to cluster major lineages and ecological origins of SE.

Significance: Our results suggested that whole genome sequencing (WGS) based approaches make a viable platform for the evaluation and benchmarking of other molecular subtyping methods.