The listeria outbreak that killed three and prompted Texas ice cream company Blue Bell Creameries to recall every one of its products late last month is the latest example of how genetic epidemiology is changing the detection of foodborne illnesses.

Two years ago, the U.S. Centers for Disease Control and Prevention (CDC) in Atlanta launched a pilot program to sequence the DNA of every listeria sample tied to an illness in the United States—all told, about 800 per year.

“Now that we’re turning whole-genome sequencing on, we’re identifying outbreak after outbreak,” says Brendan Jackson, a medical epidemiologist with CDC. “We’re also finding smaller outbreaks that we weren’t able to find before.” They’re also finding them originating in previously unsuspected foods, from caramel apples to ice cream.

The new detection method can identify gaps in the food safety system, especially when used alongside similar efforts by the Food and Drug Administration (FDA) to sequence samples from food and from the places where food is prepared, from factories to distribution centers, Jackson says. 

Whole-genome sequencing allows scientists to accurately compare every single DNA base pair in samples, giving them “a much sharper look at the differences and similarities in the strains,” Jackson says. The process takes longer—72 hours for testing compared with 48 hours—and it costs more. CDC’s goal is to create a national DNA database for foodborne pathogens from clinical samples that could be integrated with an already-existing FDA database of foodborne pathogens from food and environmental samples. The agency plans to add other foodborne illnesses, including salmonella and the most common disease-causing strain of Escherichia coli, to the project within 3 years.