Our genomes, which include all of our genetic material, are huge. If you started reading a genome sequence aloud with all the letters representing DNA bases in the genetic code ("C, A, T, G, G, T, C, G, A," etc.) and didn't stop, it would take you about nine-and-a-half years to reach the end, 3.2 billion letters later.

But what do all those letters mean, and how do our genomes affect our health? Our bioinformatics faculty are in the business of figuring that out. It's like decoding a language starting with just the letters.

In a paper published in PLoS Computational Biology, Assistant Professor Matt Hibbs, Ph.D., demonstrates that it's possible to predict genes associated with a particular disease, in this case osteoporosis, using existing data. He used large datasets—essentially equivalent to huge collections of letters and sometimes sentence fragments—and applied computing algorithms of his design to find patterns and identify gene function. He then tested and confirmed his predictions using traditional experimental methods.

"Two of the genes we experimentally confirmed were not even candidates in previous studies," says Hibbs. "The results show that we can apply our methodology to identify gene networks—multiple genes that work together in the same process or pathway—and pinpoint individual genes associated with disease." Hibbs' work will greatly speed gene discovery and add meaning to the language of life and health that we're only beginning to understand.

How have gene discoveries affected the research of other diseases? Learn more