Finding Genomic Ontology Terms in Text Using Evidence Content
Author Information
Author(s): Francisco M Couto, Mário J Silva, Pedro M Coutinho
Primary Institution: Departamento de Informática, Faculdade de Ciências da Universidade de Lisboa, Portugal
Hypothesis
Can an unsupervised method effectively identify biological properties in unstructured text using evidence content?
Conclusion
An automatic annotation system can effectively use our method to identify biological properties in unstructured text.
Supporting Evidence
- FiGO achieved a good performance when compared to other submissions in BioCreative tasks.
- The method identified Gene Ontology annotations and their evidence in a set of articles.
- FiGO does not require training data, making it efficient for recognizing properties.
Takeaway
The study created a tool that helps find important biological terms in scientific texts by looking at how often certain words are used.
Methodology
The study introduced FiGO, an unsupervised method that identifies biological properties in text based on the evidence content of their names.
Potential Biases
The method may have biases due to its reliance on exact matching and not considering context.
Limitations
The method predicted obsolete GO terms and did not filter terms that could not be annotated with human proteins.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website