Wednesday, May 2, 2012

How #BigData is Fighting #MultipleSclerosis @SUNYBuffalo

SUNY Buffalo researchers will use IBM Netezza appliance and third-party software to seek cures for multiple sclerosis.
By Ken Terry
This article was originally published at InformationWeek.com on May 1st, 2012.

Researchers at the State University of New York (SUNY) at Buffalo are harnessing "big data" technology from IBM and another firm in the fight against multiple sclerosis.

The scientists will use IBM's Netezza analytics appliance, combined with software from Revolution Analytics, to analyze huge databases and seek correlations among genetic, clinical, and environmental factors that might help reveal the causes or the acuity of MS in particular individuals. AdTech Ad

The insights that the researchers derive from this approach might help pharmaceutical companies develop new drugs or might help physicians and patients better manage the disease, said Shawn Dolley, VP of big data, healthcare, and life sciences for IBM, in an interview with InformationWeek Healthcare.

The SUNY Buffalo scientists have been using Netezza, a massively parallel computing system, for the past two years and have published some papers based on that research, Dolley said. But the addition of Revolution Analytics' application will greatly increase the number of variables they're able to include in their analyses, he said.

The big difference between what the researchers were doing earlier and what they're able to do now is that, instead of using computers to test a particular hypothesis, they will use the Big Data approach to spread as wide a net as possible.

"The SUNY Buffalo people have said with the decreased cost of the IBM Netezza and its amazing processing power, they're going to let the computer discover what the relevant phenotypes are and look at every combination of every possible variable across a large number of patients," Dolley said. Phenotype refers to the physical characteristics of a patient that result from the interaction of his genetic makeup with the environment. According to an IBM press release, the researchers will be able to study more than 2,000 genetic and environmental factors that might contribute to MS. The reason for this scattershot approach is that scientists still understand relatively little about this condition, which affects 400,000 people in the U.S. MS is now regarded as a type of auto-immune disease, but environmental, genetic, and infectious factors might play roles in MS.

The UB researchers will incorporate patient data including medical records, lab results, MRI scans, and patient surveys, as well as genomic datasets obtained from the National Institutes of Health and other sources. Among the specific factors to be examined are patients' gender, geography, ethnicity, diet, exercise, sun exposure, and living and working conditions.

Unlike Netezza's famous cousin Watson, which is designed to work with unstructured data, Netezza needs structured databases to do its magic, Dolley noted. Genetic data and clinical documentation in the free text portions of electronic health records are unstructured. So the scientists will have to spend a considerable amount of time cleaning up their data and making it manageable. But once the information is ready, IBM's system can analyze it all within minutes.

"The payoff for us is to find something new: a strain of MS or some phenotype variable that says something about the patients, such as whether these folks will respond to a particular type of treatment that will emerge as a set of clinical trials or drug discovery," said Dolley.

Commenting on the Big Data approach in the IBM announcement, Dr. Murali Ramanathan, the lead researcher at SUNY Buffalo, said, "No two people [with MS] share the exact same symptoms, and individual symptoms can worsen unexpectedly .... Identifying common trends across massive amounts of MS data is a monumental task that is much like trying to shoot a speeding bullet out of the sky with another bullet. IBM analytics helps our researchers fine tune their aim and match the speed of analysis with the rate of data coming into our systems. Our goal is to demystify why the disease progresses more rapidly in some patients and get those insights back to other researchers, so they can find new treatments."

No comments:

Post a Comment