NEW DATA-MINING EPIDEMIOLOGY COULD SPEED DISEASE CURES

Posted on November 17, 2010 by Jeffrey Newman

A couple of weeks ago, I wrote about Sergey Brin’s efforts to speed up the scientific research to find a cure for Parkinson’s disease by using data-mining through large numbers of volunteer members of social media networks. Brin, whose mother has Parkinson’s wants to break free of the “classic” scientific method: hypothesis, analysis, peer review publication. He seeks to collect information first and search for patterns via the use of algorythm computing power. This article lead me to search for similar efforts.  I found a project using data mining from a large population, also involving Parkinson’s Disease. It is a collaboration between investigators at the MassGeneral Institute for Neurodegenerative Disease (MIND) and the Harvard School of Public Health. Researchers are actually using tissues, bloods and data gathered in previous clinic studies of large populations. The participants, in their 50’s, 60’s and 70’s already have the disease and the team compares them to those patients who did not develop Parkinson’s, examining what they ate and drank, where they lived and what vitamins they tooks and what was measured in their blood. The goal was to find new leads to how Parkinson’s is triggered or prevented. This retrospective study found a link between the blood levels of a natural metabolite, urate (uric acid) and a reduced risk of developing Parkinson’s. In addition, by examining large clinical studies in which Parkinson’s had been tracked for years, where the urate was incidentally measured in more than 1600 patients, the team discovered that higher brain or blood levels of urate predicts a substantially slower rate of progression of the disease.

The Team Leaders are Dr. Michael Schwarzchild at MIND ad Dr. Alberto Ascherio at the Harvard School of Public Health.

Obviously, there is a long way to go on this research, but studies are under way with Parkinson’s patients see what affect an increase in urate might have on the progression of the disease. It is very interesting work, not only for what it could do for this degenerative disease but also what it could mean for the potential for this kind of analysis in the future to sift out leads for medical researchers, in order to truncate the process of developing therapeutics. It seems to me that a critical question is whether the methodology of data mining to seek trends, correlations and differences may be tapping into the biological processes which have developed over thousands of years and as such might reveal a series of pathfinders for our scientists. If results are exacted from such methods, it would seem logical to place significant resouces into larger computer driven projects with the thought of disease prevention and suffering prevention as well as alleviating some of the economic burden of treating diseases like cancer. IF ANYONE IS AWARE OF SIMILAR STUDIES OR PLANS FOR SUCH STUDIES, PLEASE COMMENT HERE AS I AM TRYING TO LEARN AS MUCH AS POSSIBLE ABOUT THIS PHENOMINON AND ITS STRENGTHS AND WEAKNESSES. Thanks, Jeff