Frequently Asked Questions

Q: What is the purpose of this resource?

A: To provide a portal to the entire biomedical literature focussed on SARS-CoV-2, MERS-CoV and SARS-CoV. It categorises papers into a set of categories to make it easier to find relevant research and indexes by important concepts (e.g. drugs, proteins, locations, etc).

Q: How do I find papers that mention a specific therapeutic?

A: We find drug names extracted from Wikidata in papers and use a set of synonyms to extract them.

Q: What machine learning methods are used in this resource?

A: We use a deep learning (BERT-based) supervised machine learning to identify the categories of the papers. A set of papers has been annotated with relevant categories and used as training data for a system that uses the text of the title and abstract to make predictions. This is complemented with heuristics to identify specific categories, e.g. clinical trial registration numbers.

Q: Are all the papers manually curated?

A: No. A small number have been manually curated and machine learning has been used to predict the categories for all other papers

Q: Is this up-to-date?

A: Yes. The latest version of PubMed and the CORD-19 dataset are downloaded and processed every day.

Q: Can I download or export the data?

A: Yes. Each page with a table has an export button in the top right which allows you to download the table data as a CSV or JSON file. Alternatively, you can download the entire CoronaCentral dataset from Zenodo. This is regularly updated.

Q: I've found a mistake with a paper. What should I do?

A: Please flag the paper using the Flag Mistake button on the paper or use the Feedback page.

Q: Who created this resource?

A: This was created by Jake Lever, during his postdoctoral work supervised by Russ Altman in the Helix Group at Stanford University. He is now a lecturer in the School of Computer Science at the University of Glasgow

Q: How is this research funded?

A: This project has been funded through the Chan Zuckerberg Biohub and through the National Library of Medicine LM05652 grant.

Q: Where can I get more details?

A: This work has been published in PNAS. The preprint is also available at bioRxiv. The code for the analysis also available at https://github.com/jakelever/corona-ml.

Q: How do I cite this research?

A: Please cite the PNAS paper. Below is some Bibtex if that's helpful.

@article {coronacentral,
   author = {Lever, Jake and Altman, Russ B.},
   title = {Analyzing the vast coronavirus literature with {C}orona{C}entral},
   volume = {118},
   number = {23},
   elocation-id = {e2100766118},
   year = {2021},
   doi = {10.1073/pnas.2100766118},
   publisher = {National Academy of Sciences},
   issn = {0027-8424},
   URL = {https://www.pnas.org/content/118/23/e2100766118},
   eprint = {https://www.pnas.org/content/118/23/e2100766118.full.pdf},
   journal = {Proceedings of the National Academy of Sciences}
}

Q: Who should I contact to ask a question?

A: Please contact Jake Lever at jlever@stanford.edu

Q: What other projects and resources does this website rely upon?

A: This website is built using NextJS and uses the SB Admin 2 MIT-licensed Bootstrap template. This website makes use of FontAwesome assets under a Creative Commons Attribution 4.0 International license. It uses ChartJS with react-chartjs-2 to render charts. It uses React Data Table for interactive tables. It also uses Leaflet with React Leaflet to render maps from OpenStreetMap. The code for the website also available at https://github.com/jakelever/corona-web.