Automated Metadata Extraction for Influenza A Virus Analysis
Author Information
Author(s): Olivo Miotto, Tin Wee Tan, Vladimir Brusic
Primary Institution: Institute of Systems Science, National University of Singapore
Hypothesis
Can automated metadata extraction improve the quality of large-scale protein sequence analysis of influenza A viruses?
Conclusion
Automated knowledge aggregation can significantly reduce manual corrections needed for large-scale analyses of influenza A virus sequences.
Supporting Evidence
- Over 40,000 annotated Influenza A protein sequences were collected.
- Metadata values were automatically extracted with accuracy exceeding 96% in most cases.
- The study identified 7640 isolates through automated methods.
Takeaway
This study shows that using computers to gather and clean data about viruses can help scientists work faster and more accurately.
Methodology
The study used structural and semantic rules to automatically extract and clean metadata from over 90,000 influenza A records.
Potential Biases
There is a risk of bias if records from smaller projects are discarded due to lower metadata quality.
Limitations
The study faced challenges with semantic heterogeneity and inconsistencies in public databases.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website