Bio Data-Cruncher Hits Jackpot
By Kari Lynn Dean
02:00 AM Oct. 28, 2004 PT
While surfing the internet not long ago, a Harvard biologist stumbled upon a pile of public research data including unpublished leftovers from an unresolved genetics study. It wasn't unusual: Data like that can be found all over the web.
But then 33-year-old biology professor Vamsi Mootha did something remarkable. Using a unique computational method, he mined the data and identified a gene underlying a rare but fatal pediatric disorder called Leigh syndrome, French-Canadian variant, or LSFC. Astonishingly, he did it in a single weekend.
. . .
A relatively new technology, called microarrays, enables interrogation of every gene to determine how active it is. Rather than just look at a slice of diseased cell tissue under a microscope, scientists can see how many of the 23,000 genes are switched on or off and to what degree. Multiply all that data by all the patients in research studies such as Mootha encountered, and the result is an intimidating mass of numbers to crunch and assess.
"Vamsi got a hold of the data from the internet, but he said you can't compare gene by gene. You'd be doing so many comparisons, you aren't going to find anything statistically significant," said Alan Attie, a University of Wisconsin biochemistry professor.
"Vamsi re-curated the list of genes, making about 120 categories by functional group," such as genes that make fat or carbohydrates, or control respiration, etc., Attie said. "It turned out that the mitochondrial respiration group showed a big difference. But looking at the individual gene level, there would have been only modest differences."