ScholarMuse DocsHow to cite ScholarMuse data
Cite the original source, not ScholarMuse. Every record carries the source, link, licence, attribution and retrieval time you need.
ScholarMuse fetches data from public research sources and passes it on. The data belongs to those sources, so cite the original source, the same way you would if you had looked it up yourself.
What each record gives you
Every record your agent receives carries:
| Field | Use it for |
|---|---|
source | Which database the record came from |
url | The record's own page at the source, when there is one |
license | The terms the source publishes its data under |
attribution | The credit line the source asks for |
retrievedAt | The date and time the record was fetched; use it as your access date |
The facts themselves, such as the DOI, accession number or series ID, are in the record's data. See What your agent gets back.
Examples
A paper. Cite the paper itself, with its authors, title, journal, year and DOI. It does not matter that your agent found it through OpenAlex or PubMed. Before you cite it, open the DOI link and check the details, or ask your agent to verify your reference list.
A data series. Cite the provider and the series, with your access date. For an unemployment series from FRED:
U.S. Bureau of Labor Statistics, Unemployment Rate [UNRATE], retrieved from FRED, Federal Reserve Bank of St. Louis, https://fred.stlouisfed.org/series/UNRATE, accessed 12 October 2026.
Take the access date from retrievedAt, and follow the attribution line on the record. Some FRED series are copyrighted by their original provider, which the record's licence will say.
A database entry. For a gene, protein, compound or structure, cite the database and the identifier, for example the UniProt accession or PDB ID, plus the database's own reference paper if your journal expects one. The attribution line tells you who to credit, for example "Protein data from UniProt (uniprot.org), UniProt Consortium, CC BY 4.0."
A dataset. Cite the dataset by its DOI, as listed in the record. Datasets found through DataCite, Zenodo or Dryad usually have one.
Asking your agent
Because the source travels with each record, you can ask:
- "Which database did each of these numbers come from?"
- "Add a data sources section with the attribution lines and access dates for everything you used."
- "Format these papers as references in APA style, using the DOIs from the records."
Check the formatted references yourself. The records come from the sources; the formatting is the agent's work.
Do I need to cite ScholarMuse?
No. ScholarMuse is the route to the data, like a library catalogue. If your journal asks you to describe how data was collected, you can write in your methods that records were retrieved through ScholarMuse (scholarmuse.ai) from the named sources on the given dates.
Licences
Each source has its own licence. Many are open (for example CC0 or CC BY 4.0); some have terms of use. Follow the licence and attribution on each record when you publish or share the data. The Sources page lists every source with its licence.
Figures drawn by ScholarMuse are made by an AI model, not taken from a source. Follow your journal's rules on AI-generated images. See Scientific figures.