More about
Dr. Wren:

Dr. Wren's profile

Dr. Wren 101
(for non-scientists)

Dr. Wren's CV in brief

Recent Publications

Full Publications List

Research

Professional Affiliations

Past Projects

Recent and Ongoing Projects

Patents Received & Pending

Arthritis and Immunology Research Program

Dr. Wren In The News

OMRF receives $26 million for two federal research grants

Jonathan D. Wren, Ph.D. Assistant Member, Arthritis and Immunology Research Program

Recent and Ongoing Projects
Text Mining/Knowledge Discovery
A fair amount of my current research revolves around methods of discovering new knowledge using large-scale literature analysis. Much of this work is embodied in a software package called IRIDESCENT. IRIDESCENT was written for the assimilation and analysis of unstructured literature domains for the purpose of discovering new knowledge implicit from existing information. It has been very useful so far in a number of different areas including knowledge discovery, ontology construction and microarray analysis. I continue to work on improving, refining and extending methods of inference and identifying commonalities. I am also working on information retrieval and data classification methods for fully or semi-automated database construction, currently with regards to short sequence related databases. These databases currently require a lot of manual labor which can be time-consuming and expensive, not to mention generally tedious and sometimes inconsistent.

Networked Information Resources
Clearly, the Internet age is changing the way we conduct research via the increasing availability of literature, software and analysis tools online. Unfortunately, the Internet is also dynamic, meaning that the continued availability of electronic resources is not necessarily guaranteed. I've conducted two studies of URLs published in MEDLINE abstracts and found that, just as in other fields, their availability decays over time. For example, 93% of 2002 URLs were still up while only 42% of 1995 URLs were. For most authors, though, URL decay is out of their control. In a similar vein, I studied the decay rates of corresponding author emails published in MEDLINE and find a similar trend, except emails decay much faster. I have also studied the online availability of information and found (using an API from Google) that the probability a journal article can be found online at a non-journal website rises with the impact factor of the publishing journal and recency of the paper.

Microarray Analysis
Ah, microarrays. High-throughput technologies aren't changing the way we do science, but the scale by which we do it. Each array run is usually for a specific purpose (e.g., to compare normal and cancerous breast tissue samples), yet the data gathered is genome-wide. I'm working on meta-analytic methods of this data to identify regulatory trends as well as trying to integrate genes with similar patterns of behavior under heterogeneous conditions with a literature analysis to identify commonalities. This is just now passing out of its nascent stage and we (at OMRF) have tested several of these genes and validated experimentally that their predicted functions are real (manuscript in preparation).

Mailing Address
Arthritis and Immunology Research Program, MS 58
Oklahoma Medical Research Foundation
825 N.E. 13th Street
Oklahoma City, Oklahoma 73104

Contact Information
Phone: (405) 271-6989
Fax: (405) 271-4110