![]() |
|||
|
Research | Core Facilities | Patient Studies | Tech Transfer | Seminars | Intranet | Careers | Search | Contact Us | Ways To Give HOME |
|||
|
More about Arthritis and Immunology Research Program
|
Research Interests Because the relevant findings of scientific research are almost always summarized in published form and increasingly available electronically, part of my work focuses on developing new methods of text mining, or data-mining using the scientific literature as a data source, to model this data. Computers can easily process 16 million published papers but cannot understand implications or weigh evidence in ways that humans can. Humans, however, have neither the time nor interest to read anywhere near 16 million papers and wouldn’t remember every detail even if they could. So my work in this area has focused on finding the middle ground – getting computers to do the reading and assimilate the data and/or implications in summary form for humans. One of the programs we have developed, IRIDESCENT, uses known biomedical relationships to infer what is not known but is plausible. It has so far successfully identified several novel relationships, some of which have been validated experimentally (e.g. that a drug known for its neuroactive properties could also affect the progression of a cardiac illness). Finally, because of this growing abundance of data, there are more and more findings that are not reported in the traditional journal format. Rather, they are deposited in specialized databases (DNA sequences, microarray data, protein-protein interactions, etc). Recently, we have begun to focus on ways to integrate this information into a framework that will help make sense of the diseases that OMRF is working to cure. Joined OMRF Scientific Staff in 2007. Mailing Address
|