From: Lu, Zhiyong (NIH/NLM/NCBI) [E]
Sent: Friday, May 01, 2009 10:54 AM
To: NLM/NCBI List ncbi-seminar
Subject: Special Seminar Thur May 7

Follow Up Flag: Follow up
Flag Status: Red
Special seminar by PostDoc candidate Jiao Li, at B2 library, Thursday May 7, 11am. 


How does literature mining help biomedical studies?
-- From information retrieval to disease-protein-drug association mining
Since the amount of biomedical literature is increasing at a considerable rate, it is no longer possible for researchers to keep up-to-date with all relevant literature manually. Literature mining comes to help them to find relevant articles, identify field trends, extract specific relations, and generate potential hypothesis. In my talk, I will focus on the following three aspects of biomedical literature mining in my Ph.D. studies. (1) Query expansion and document rank in biomedical literature retrieval: We participated in TREC Genomics Track from 2004 to 2007, and evaluated our methods using the standard data. (2) Ontology-based biomedical literature mining for competitive analysis: An ontology-based literature mining system was designed to support competitive analysis in the biosciences. It is able to track field development and identify experts with similar research profiles. (3) Disease-specific drug-protein connectivity maps built from molecular interaction networks and PubMed abstracts: After validating our method in Alzheimer’s Disease, breast cancer, and pancreatic cancer case studies, we implemented this integrated method into a web server. It enables biomedical researchers to investigate and evaluate comprehensive molecular associations between disease-specific proteins and candidate drugs.

Jiao Li is a Ph.D. candidate at the Department of Computer Science and Technology, Tsinghua University, China. She has been in a joint-training Ph.D. program and worked in Purdue University School of Science since Sep. 2007. Her research interests include biomedical literature mining, knowledge representation, data mining and disease-specific biomedical information integration.