TB crush

Video gamers help develop a better test for tuberculosis

Featured Media for TB crush

In the video game Eterna, players design RNA molecules to solve medical problems. Photo courtesy of Riju Das.

Purvesh Khatri, PhD, had a better test for active tuberculosis. TB is hard to diagnose accurately, but Khatri and his team had discovered that the ratio of three particular RNA molecules from a blood sample could show if a person had the disease. One thing was missing, however: a cheap way to deploy the test in some of the most remote locations in the world.

While at the Big Data in Biomedicine conference at Stanford last year, Khatri came upon a surprising solution. Sitting next to him was Rhiju Das, PhD, who told him about a research team ideally suited for the project: 100,000 registered players of Eterna, a video game Das launched five years ago to design useful biomolecules.

The two Stanford professors hatched a new challenge for the players: to design a single RNA molecule, dubbed OpenTB, that could calculate the ratio among the three RNA molecules. The two recruited assistant professor of genetics Will Greenleaf, PhD, who had just invented a technology that could test hundreds of thousands of player-designed molecules.

If the players succeed, they’ll help save millions of lives. TB infects a third of the world’s people and kills about 1.5 million each year. Yet there’s no easy-to-use blood test that can detect active infection and identify who could benefit from antibiotics. OpenTB would be the indicator molecule needed to create an assay as simple as a home pregnancy test.

Unlike two-stranded DNA, RNA is single-stranded, floppy and spontaneously folds into myriad shapes. Over the years, Eterna players have become expert in designing complex RNA molecules, says Das, an associate professor of biochemistry.

Six months into the new game, the results are encouraging, says Das. “We have promising leads, in several cases from brand-new players who started playing Eterna after hearing about the OpenTB challenge,” he says.

Although OpenTB would be one RNA molecule, it would have three parts, each of which would bind to one of the TB-related RNA molecules.

The OpenTB molecule must also assume different shapes depending on the proportions of the three kinds of RNA. “If I have a lot of RNA molecules A and B around,” says Das, “OpenTB will fold into shape 1. But if there’s a lot of C around, OpenTB will fold into shape 2.”

For OpenTB to work, shape 1 also must be able to bind to a fluorescent tag, while shape 2 must not bind to the tag. So individual molecules with shape 1 would emit light and those with shape 2 would not.

By measuring the brightness, says Das, you can calculate what proportion of Eterna molecules have folded into shape 1, revealing the proportion of RNAs A and B relative to C. If the light is above a certain threshold of brightness, you know the patient has active tuberculosis.

“I love this idea because it changes the biological research paradigm,” says Khatri, an assistant professor of medicine. Khatri came up with the idea for the TB test after analyzing gene expression data from more than 1,000 blood samples, all from publicly available data sets.

If OpenTB is successful, says Khatri, “it would allow us to say, ‘We can use publicly available data — ultimately provided by patients themselves — to find a diagnostic signature of one of the biggest killers of mankind. And then we can engage the public to design molecules that can help deploy that test using a video game platform.’”

Author headshot

Jennie Dusheck

Jennie Dusheck is a science writer for the medical school's Office of Communication & Public Affairs. Email her at dusheck@stanford.edu.

Email the author