Scaling up data curation using deep learning: An application to literature triage in genomic variation resources.

Manually curating biomedical knowledge from publications is necessary to build a knowledge based service that provides highly precise and organized information to users. The process of retrieving relevant publications for curation, which is also known as document triage, is usually carried out by querying and reading articles in PubMed. However, this query-based method often obtains unsatisfactory precision and recall on the retrieved results, and it is difficult to manually generate optimal queries. To address this, we propose a machine-learning assisted triage method. We collect previously curated publications from two databases UniProtKB/Swiss-Prot and the NHGRI-EBI GWAS Catalog, and used them as a gold-standard dataset for training deep learning models based on convolutional neural networks. We then use the trained models to classify and rank new publications for curation. For evaluation, we apply our method to the real-world manual curation process of UniProtKB/Swiss-Prot and the GWAS Catalog. We demonstrate that our machine-assisted triage method outperforms the current query-based triage methods, improves efficiency, and enriches curated content. Our method achieves a precision 1.81 and 2.99 times higher than that obtained by the current query-based triage methods of UniProtKB/Swiss-Prot and the GWAS Catalog, respectively, without compromising recall. In fact, our method retrieves many additional relevant publications that the query-based method of UniProtKB/Swiss-Prot could not find. As these results show, our machine learning-based method can make the triage process more efficient and is being implemented in production so that human curators can focus on more challenging tasks to improve the quality of knowledge bases.

Cardiovascular imaging research priorities.

OBJECTIVES: Two interlinked surveys were organised by the British Heart Foundation Data Science Centre, which aimed to establish national priorities for cardiovascular imaging research. METHODS: First a single time point public survey explored their views of cardiovascular imaging research. Subsequently, a three-phase modified Delphi prioritisation exercise was performed by researchers and healthcare professionals. Research questions were submitted by a diverse range of stakeholders to the question 'What are the most important research questions that cardiovascular imaging should be used to address?'. Of these, 100 research questions were prioritised based on their positive impact for patients. The 32 highest rated questions were further prioritised based on three domains: positive impact for patients, potential to reduce inequalities in healthcare and ability to be implemented into UK healthcare practice in a timely manner. RESULTS: The public survey was completed by 354 individuals, with the highest rated areas relating to improving treatment, quality of life and diagnosis. In the second survey, 506 research questions were submitted by diverse stakeholders. Prioritisation was performed by 90 researchers or healthcare professionals in the first round and 64 in the second round. The highest rated questions were 'How do we ensure patients have equal access to cardiovascular imaging when it is needed?' and 'How can we use cardiovascular imaging to avoid invasive procedures'. There was general agreement between healthcare professionals and researchers regarding priorities for the positive impact for patients and least agreement for their ability to be implemented into UK healthcare practice in a timely manner. There was broad overlap between the prioritised research questions and the results of the public survey. CONCLUSIONS: We have identified priorities for cardiovascular imaging research, incorporating the views of diverse stakeholders. These priorities will be useful for researchers, funders and other organisations planning future research.