Increasing metadata coverage of SRA BioSample entries using deep learning-based named entity recognition

Keyphrases

• Deep Learning Methods 100%
• Named Entity Recognition 100%
• Sequence Read Archive 100%
• Biosamples 100%
• Recurrent Neural Network 50%
• Metadata Annotation 50%
• Meta-analysis 25%
• Public Repositories 25%
• Withholding 25%
• Manual Curation 25%
• National Center for Biotechnology Information 25%
• Prediction Accuracy 25%
• Area under the Receiver Operating Characteristic Curve 25%
• Overall Accuracy 25%
• Training Model 25%
• Reproducible Research 25%
• Evaluating Performance 25%
• Information Sequence 25%
• Multiple Issues 25%
• Attribute-value Pair 25%
• Recognition-based 25%
• Quality Metadata 25%

Computer Science

• Named Entity Recognition 100%
• Deep Learning Method 100%
• Annotation 66%
• Recurrent Neural Network 66%
• Reproducibility 33%
• Public Repository 33%
• Characteristic Curve 33%
• Training Model 33%
• Prediction Accuracy 33%
• Information Sequence 33%
• Attribute Value 33%
• Disease Condition 33%