Salmonella Serotyping Through Whole Genome Sequencing

November 2014

Salmonella is the most prevalent bacterial foodborne pathogen in the United States, causing more than 1 million cases of illness annually, and is the largest economic burden among all bacterial pathogens. For decades, the U.S. National Salmonella Surveillance System has been based on serotyping, a subtyping method relying on the agglutination of Salmonella cells with antisera. There are now more than 2,600 distinct Salmonella serotypes reported worldwide, all of which can potentially cause human gastroenteritis.

While still playing a vital role in foodborne salmonellosis surveillance and outbreak investigations, serological identification of Salmonella serotypes is time-consuming (more than 2 days to complete), logistically demanding (production, quality control, and maintenance of hundreds of antisera for full-set serotyping), and sometimes imprecise. In addition, the ambiguities and idiosyncrasies associated with phenotypic characterization of cell surface antigens complicate identification, reporting, and surveillance of Salmonella, making it highly desirable to have a genotypic alternative.

Routine and real-time implementation of whole genome sequencing (WGS) is poised to transform public health microbiology, as evidenced by recent ongoing initiatives at federal agencies, including FDA’s GenomeTrakr and CDC’s Advanced Molecular Detection. By making the entirety of genetic information readily available, WGS not only allows closely related isolates to be differentiated with unparalleled resolution but also promises the integration of a multitude of identification, subtyping, and characterization workflows into a single, comprehensive, and highly efficient platform.

Funded in part by contributions from CFS Board of Advisors and in collaboration with the Enteric Disease Laboratory Branch at CDC, we recently developed “SeqSero”, a bioinformatics tool that allows rapid (within minutes), accurate, and near full-spectrum serotype determination of Salmonella. More than 4,000 isolates representing more than 200 different serotypes have been tested using SeqSero, achieving more than 95% accuracy as compared to about 90% with traditional subtyping.

We have recently launched a cloud-based web application ( pending a domain name) for public access to SeqSero for beta testing. The application runs on all major web explorers and mobile devices and requires no prerequisite bioinformatics skills for analysis. We anticipate the release of SeqSero to the public by the end of 2014 after rigorous testing by federal and state public health agencies, and academic partners.