Related Experiment Videos
Knowledge discovery by automated identification and ranking of implicit relationships
Jonathan D Wren1, Raffi Bekeredjian, Jelena A Stewart
1Advanced Center for Genome Technology, Department of Botany and Microbiology, The University of Oklahoma, 620 Parrington Oval Rm. 106, Norman, OK 73019, USA. Jonathan.Wren@ou.edu
Bioinformatics (Oxford, England)
|February 13, 2004
Summary
This study introduces a computational method to discover novel biomedical relationships from scientific literature. The approach successfully identified chlorpromazine as a compound that may reduce cardiac hypertrophy progression.
Area of Science:
- Biomedical Informatics
- Computational Biology
- Data Mining
Background:
- The rapid growth of scientific literature makes manual identification of implicit relationships challenging.
- Discovering novel connections between biomedical entities requires advanced analytical tools.
Purpose of the Study:
- To develop and validate a computational method for identifying and ranking implicit relationships within scientific reports.
- To enable large-scale analysis of potential connections between unrelated biomedical concepts.
Main Methods:
- Constructed a network of tentative relationships based on co-occurrences of biomedical entities (genes, diseases, chemicals) in MEDLINE records.
- Ranked shared relationships between unrelated objects against a random network model to assess statistical significance.
- Validated the method by comparing rankings with known relationships and predicting novel ones.
Main Results:
- The ranking method correlated with object co-occurrence probability and frequency, indicating its suitability for discovering novel relationships.
- The approach identified compounds potentially affecting cardiac hypertrophy.
- Laboratory testing confirmed that chlorpromazine reduced cardiac hypertrophy progression in a rodent model.
Conclusions:
- The developed computational method effectively identifies and ranks potential novel relationships in biomedical literature.
- This approach facilitates the discovery of new therapeutic targets and drug candidates, as exemplified by the cardiac hypertrophy findings.