Jonathan Usuka is driving his computer, scanning a genome database for irregular areas of DNA that could be responsible for causing asthma. With a couple of clicks, images of four of the database’s 19 chromosomes begin to emerge on the screen.
“We know that one set of researchers spent two years identifying chromosomes 2 and 7, and another set took two years to identify chromosomes 10 and 11,” says Usuka, as the four chromosomes take the shape of rectangular, overlapping smokestacks. “And you and I have just spent about 20 seconds identifying all four—2, 7, 10 and 11. That ’s the big deal here.”
Working with colleagues at Roche, a pharmaceutical company where he is a part-time consultant, Usuka, a doctoral candidate in chemistry, compiled a database of the 15 most commonly used strains of lab mice. He then developed software that scans for irregularities in mouse dna, which is 80 percent similar to humans’, in an attempt to speed up the search for genes linked to diseases. In June, he and eight co-authors published a paper in the journal Science describing the program, Digital Disease.
“I never touch mice and I’m not thrilled about the whole animal testing thing,” says Usuka, who has nonetheless developed a particular fondness for C57Bl6J, an intelligent, long-living black mouse that likes to guzzle alcohol. Instead, Usuka does computational chemistry, testing his software against known genetic discoveries. So far.
“But in a year or two, instead of just reproducing other people’s results quicker, we are going to do something new and real, and do it computationally,” he says. “We will take a disease that we don’t know the genes or locations for, and we’ll find those genes and locations. And then we’ll have drug targets, and then we’ll cure things.”
