University of Pittsburgh gets wired for speed with Apple Xserve G5 cluster


PITTSBURGH, April 13 Every week in the CBS network's new hit series Numb3rs, an FBI agent relies on his math genius brother to find patterns and equations that help to solve crimes. With its new Apple Xserve G5 computing cluster, the University of Pittsburgh's Graduate School of Public Health (GSPH) is solving double-helix puzzles in human genetics every day and faster than a speeding FBI-issue bullet.

Using the school's newly installed 125-node Xserve cluster, more than 30 investigators and scientific teams, engaged in more than 120 complex research projects have all the computing power they could ask for, according to M. Michael Barmada, Ph.D., associate professor in the department of human genetics at GSPH. Paid for with a shared resource grant from the National Institutes of Health, the human genetics computing cluster is among the fastest at an academic medical center in the United States.

"Our division of statistical genetics is looking for genes that influence diseases," explained Dr. Barmada, whose research focuses on the genetic epidemiology of common yet complex disorders such as diabetes and inflammatory bowel disease. "In a sense, we're gene hunters."

Dr. Barmada and his colleagues are using their new computing power dubbed locally as the Gattaca Cluster for the 1997 feature film to analyze data involving the many genes that lead to variations in human traits, from those that regulate differences in height and bone density to those that influence susceptibility to disease. The Gattaca Cluster supports dozens of highly specialized statistical genetics applications used in ongoing research projects. Using genetic analysis algorithms, the computer can evaluate unique markers that characterize a segment of an individual's genetic material and attempt to correlate the markers with patterns of transmission through a family or population.

"Our studies even those that analyze just 400 markers in a population of, say, 20 families with 25 people each can be extremely compute intensive," said Dr. Barmada.

David Whitcomb, M.D., Ph.D., professor of medicine, cell biology and physiology and human genetics at the University of Pittsburgh School of Medicine, has done extensive studies on the genetic basis of pancreatic cancer, the fifth leading cause of cancer death in the United States. He is excited about the human genetics cluster's capacity to increase productivity on these studies, as well as others he is pursuing on pancreatitis recurrent inflammation of the pancreas that can lead to pancreatic cancer.

"Most disease is caused by a combination of genetic mutations and environmental stress," said Dr. Whitcomb, who also is chief of the division of gastroenterology, hepatology and nutrition and director of the Center for Genomic Studies. "Such a combination might give an individual 500 times the average risk of a given disease in the general population or more."

Trying to sort out genuine risk from arbitrary observation stretches the limits of old statistical methods, Dr. Whitcomb added.

"We are looking at the possible interactions of multiple genes and environmental factors in our group of patients," he said. "There are thousands of possible genetic and environmental factors that might combine to result in disease. The computer can help to predict which factors are disease-causing and which are innocent bystanders."

This is done by simulating an entire population of people and "testing" a group of imaginary patients with laboratory results from actual patients tens of thousands of times, explained Dr. Whitcomb. Eventually, the computer is able to sort out the genetic and environmental factors that may cause complex diseases.

"With older computers, these complex calculations can take weeks," he said. "Insights into the factors that contribute to diseases give us targets for effective therapy. The new Gattaca Cluster is a terrific tool to help the discovery process."

Future plans exist to extend Gattaca's workload even more by analyzing data from microarrays and large SNP-chip studies, which can involve as many as 100,000 markers.

"This will greatly increase the number of computations required for a genetics project, but Gattaca has the horsepower and storage capacity to handle whatever we throw at it," said Dr. Barmada.

Source: Eurekalert & others

Last reviewed: By John M. Grohol, Psy.D. on 21 Feb 2009
    Published on All rights reserved.