Bioinformatics: adding computational muscle to food and water safety

It turns out there really is such a thing as information overload. As Dr. Gary Van Domselaar explains, "In genomics, our ability to generate information was exceeding our ability to analyze that information—that's why we need bioinformatics."

As for applying this new scientific discipline to food safety, Dr. Van Domselaar—chief of bioinformatics at the Public Health Agency of Canada—recalls the deadly outbreak of listeriosis in Canada in 2008. "People in different parts of the country were getting sick with an illness that we know can be transmitted through contaminated food," says Dr. Van Domselaar. "In that situation, you want to determine as quickly as possible whether you're dealing with a series of isolated, unrelated cases or a genuine outbreak that would require warnings to be issued and other steps taken to find and deal with the source of the contamination."

Enter genomics

Until relatively recently, authorities relied on so-called 'shoe-leather epidemiology' to make that determination—investigators talking to people who'd gotten sick, asking them what they ate, when and where they ate it, who they'd been in contact with and so on.

The science of genomics provided epidemiologists with a new tool. The ability to see the bacteria's DNA gave investigators a much more accurate way to tell whether the same or different strains of a bacteria were responsible for the infections—although, since the testing method sampled only a few sections of the bacteria's DNA, there was still significant room for error.

Next-generation sequencing

At the time of the listeriosis outbreak, next-generation sequencing (NGS) technology was already widely available, making an even more accurate test possible. "With NGS, virtually the entire genome of a bacterium like Listeria could be sequenced in three days," says Dr. Van Domselaar. "It allows you to see practically every one of the millions of base pairs that make up its DNA."

What was not available was a means to interpret that amount of data in a timely way. "It was possible to make a much more accurate identification of the bacterial strains, but it couldn't be done quickly enough to be useful," says Dr. Van Domselaar. "It took three days to collect the data and three months to interpret it, so investigators had to rely on the older, less accurate test method."

Addressing the high cost of foodborne illness

The 2008 listeriosis outbreak caused 24 deaths and cost the Canadian economy an estimated $240 million. Recognizing the role genomics could play in reducing the incidence and severity of such outbreaks, the Government of Canada—through the Genomics Research and Development Initiative—provided funding for the multi-year, multi-disciplinary Food and Water Safety (FWS) research project.

While some FWS researchers dedicated themselves to sequencing the genomes of hundreds of strains of Salmonella and other bacterial contaminants, Dr. Van Domselaar was charged with developing the bioinformatics platform needed to use these massive amounts of genomic data as effectively and efficiently as possible.

In collaboration with other researchers, end-users and stakeholders—and with additional funding from Genome Canada—Dr. Van Domselaar and his team have developed a bioinformatics platform called IRIDA, now being deployed across Canada.

Powerful and flexible

"IRIDA stands for Integrated Rapid Infectious Disease Analysis," says Dr. Van Domselaar. "It adds a new dimension to PulseNet—Canada's existing national network for foodborne disease surveillance—giving provincial health authorities access to the bioinformatics capacity they need to store, manage, share and analyse all kinds of genomic data."

Professor Fiona Brinkman of Simon Fraser University—a senior investigator on the project—says one of IRIDA's most important features is its adaptability. "Its modular structure allows us to easily incorporate new analysis methods, or combinations of methods, as they’re developed," explains Dr. Brinkman, "It's a platform with a lot of computational muscle but also flexibility."

Another senior investigator, Dr. William Hsiao of the University of British Columbia and the BC Centre for Disease Control, says that means IRIDA can do more than enhance food safety. "It will require additional funding, but IRIDA can be adapted to tracking any kind of infection," says Dr. Hsiao, "From flu viruses to the antibiotic-resistant bacteria that are a major concern for hospitals."

That adaptability is also developing an international dimension. Dr. Van Domselaar reports a growing number of countries are looking at the IRIDA platform to meet their bioinformatics needs.

And what about speed?

"Oh yes," says Dr. Van Domselaar. "That genomics data that would have taken three months to analyze back in 2008? We can do that more-or-less in real-time now."