IBM, Mars see gene sequencing as key to protecting the global food supply

One in six people in the United States contract a food-borne illness every year, and thousands of them die. IBM and food giant Mars think genomics may help save many of those lives, as well as billions of dollars a year.

Today, the two giant companies announced the formation of the Sequencing the Food Supply Chain Consortium. While the effort has a long and unwieldy name, and will embark on some complex science, the goal is simple: develop better knowledge about what makes food safe and unsafe.

Every year, according to IBM, food-borne disease kills 3,000 people in the U.S. In addition, Big Blue said, more than 128,000 people end up in the hospital, incurring $9 billion in medical costs. At the same time, $75 billion in tainted food is thrown away every year.

Add it all up, and the costs of contaminated food are tremendous — and that’s just in the United States.

According to James Kaufman, the manager of public health research at IBM’s Almaden Research Institute in San Jose, Calif., the new consortium was started to “demonstrate how one can use advances in microbiology and next-gen sequencing — figuring out the order of DNA or RNA, the building blocks of nature — to measure the fingerprints of the community of organisms, the bacteria, that you expect to see in food.”

The two-year joint project will begin with scientists studying the genetic marks of organisms such as bacteria, viruses, and fungi, and how they develop in a variety of food-industry environments, like ingredients, kitchens, and factories.

The idea is to understand which organisms are meant to be in food, and which are out of place. Detection is key. Know what’s not supposed to be there, and you may be on the way to making food safer. Yet, while you may think that bacteria in food is always a problem, Kaufman would say that some foods, like yogurt, can’t exist without living organisms.

The food industry has long had a detection system called PulseNet, run by the Center for Disease Control and Prevention, that “compares the ‘DNA fingerprints’ of bacteria from patients to find clusters of disease that might represent unrecognized outbreaks.”

But Kaufman said that the problem with PulseNet is that it is dependent on looking for known organisms. What if something new enters into the food system that’s never been seen before? He pointed to the melamine scandal in China in 2008, when thousands of infants were sickened, and several died, after a milk producer secretly added the ingredient — which affected the organisms in the milk — in an attempt to save money. “How do you test for” what you don’t know might be there, Kaufman said.

By sequencing the genes of organisms and creating a database of known bacteria, IBM and Mars think they can set the food industry on a path towards being able to recognize ahead of time when something is wrong with what people are eating.

In the early going, IBM will test samples from a Mars pet food factory, not once but several times over two years, aiming to generate a baseline of the known organisms in the ingredients and how they change based on the season.

Above: A sequencing readout, like those that could help companies in the food industry better understand how to detect contaminants.

Image Credit: IBM

From time to time, Kaufman added, someone will “spike the samples” with a contaminant, in order to help scientists better identify anomalies.

IBM and Mars believe that as the cost of gene sequencing and computation drop, and as the world’s food needs explode (Kaufman said the world’s food supply will need to increase by 75 percent by 2050 to keep up with population growth), companies throughout the industry could benefit by joining the consortium.

And since “we’re not going to see farms buying supercomputing clusters,” Kaufman added, IBM hopes it can generate business by making its cloud-computing tools available to the industry at an inexpensive cost.

IBM also hopes that companies in the food industry will recognize the value of being able to access a large, and growing, database of organism fingerprints. “You can always go out and hire microbiologists and computer scientists in isolation,” he said, “but you lose because you don’t have access to the larger data.”