INFORMATION IS POWER. It’s money. It’s making the world safer, smaller, and more connected. But information is nothing more than growing piles of numbers unless we’re able to interpret and understand what it means.
Underneath those mountains of data are golden nuggets of useful facts and figures waiting to be discovered. To search them out, UAB established the Knowledge Discovery and Data Mining (KDDM) Research Lab. Since 2004, an interdisciplinary team of computer scientists, engineers, physicians, and other researchers has applied the lab’s resources to a variety of applications—from monitoring and analyzing traffic patterns to identifying highlights in sports videos. Most recently, the focus of the KDDM has turned to the revolutionary field of human genome research.
The search for gene expression—finding the relatively small number of genes in a cell that are “turned on,” or capable of expressing the cell type’s unique properties—has generated a wealth of information. In the past few years, methods of surveying this biological roadmap have become much more efficient. Microarray technology, which employs glass slides containing samples of many genes arranged in a regular pattern, led the way, followed by the development of analysis software. The process continues to be time-intensive, however. “The commercial software still requires a lot of human effort to interpret the results,” says Chengcui Zhang, Ph.D., one of the KDDM’s head researchers. “Our goal is to create a fully automated system.”
Using the most powerful network of computers in the state of Alabama, Zhang and her colleagues have begun to do just that. While complete automation is still years away, they’ve been successful in reducing the valuable lab time required to perform microarray analysis—and they expect improvements to continue at a rapid pace. “This is a great application to utilize the computer cluster,” Zhang says, “but we haven’t come close to utilizing its full power yet.”
The implications of the KDDM project are significant, not only in the field of bioinformatics but also in terms of how scientists and health professionals understand fundamental aspects of human growth and development. With more advances, it’s expected that researchers will be able to uncover whole new categories of genes and deduce functionality by comparing their expression patterns with those of known genes. In addition, by enabling scientists to examine vast samples more quickly, the use of advanced microarray analysis may point to the specific genes responsible for a variety of diseases.
At the current rate of innovation in data mining, information could hold even more power and promise in coming years. Sometime in the foreseeable future, will doctors be able to ascertain your risk for every known disease—instantly—from a hair or skin sample?
“Who knows?” Zhang says. “In the future, anything is possible.”
—STEVE DUPONT