Improving Search in the OMIM Database Using Natural Language Processing
Author Information
Author(s): Cohen Raphael, Avitan Gefen, Elhadad Michael, Birk Ohad S
Primary Institution: Ben-Gurion University
Hypothesis
Can natural language processing improve the search functionality of the OMIM database?
Conclusion
The enhanced search functionality provides clinicians with an efficient tool for diagnosis and identifying similar syndromes.
Supporting Evidence
- 79,770 new connections between syndromes were identified, averaging 16 new connections per syndrome.
- Precision of predicted synonymous phrase pairs was found to be 93.5%.
- 3.2% of the phrases were recognized as negations with a precision of 89%.
Takeaway
This study made it easier for doctors to find information about genetic syndromes by improving how they search through a big database.
Methodology
Natural language processing was used to tag phrases with semantic information and cluster similar phenotypes for improved search.
Limitations
The study may not account for all variations in phenotype descriptions due to the complexity of natural language.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website