Data & Citation
Machine- and reviewer-facing facts about Oliver's mTOR
Atlas as a dataset: where it is registered, how it is identified, and how
to cite it. Everything below also appears in this dataset's
CITATION.cff and LICENSE files and in this
page's own JSON-LD, in case a script needs it rather than a human.
Registered with
| bio.tools | Oliver's mTOR Atlas is registered with bio.tools — ELIXIR's registry of bioinformatics tools and databases — listed as a Database portal under Genetics, Molecular biology and Systems biology. |
| FAIRsharing | Oliver's mTOR Atlas is registered with FAIRsharing.org (record FAIRsharing.org8905) — a curated registry of databases, standards and policies for the life sciences — listed as a Database resource under Molecular Biology, Biochemistry, Bioinformatics and Aging. |
Identifiers
| Dataset DOI | 10.5281/zenodo.22059963 — concept DOI, always resolves to the latest archived version on Zenodo |
| Version DOI | 10.5281/zenodo.23090261 — the archived v2.0.1 release of 2026-10-01 (428 studies), for citing exactly the state that release archived |
| Live dataset | 2.0.1+unreleased — the live data have changed since v2.0.1 (518 studies now, 428 in the archive) and are not yet archived; the version DOI above does not contain them |
| FAIRsharing DOI | 10.25504/FAIRsharing.691885 — resolves to the FAIRsharing registry record above |
| Curator ORCID | 0009-0008-2025-2148 (Oliver Barton) |
| License | CC BY 4.0 — free to share and adapt, including commercially, with attribution |
Sustainability and preservation
The Atlas is an independent curation project without institutional funding. Its preservation does not depend on this website staying online. Each release is deposited on Zenodo as a versioned, citable snapshot (v1.0.0, v2.0.0, v2.0.1) under a concept DOI that always points to the latest version, and the accompanying analysis is deposited as a preprint on Figshare. The full corpus is also published as open CSV and JSON files (see the downloads above) under CC BY 4.0, so anyone can keep a copy or build on it. A released version is never edited afterwards: corrections go into a new version, and every change to the method is logged with its date at /changelog/.
Live data and archived versions
Two different things carry the Atlas's name, and a citation should say which one was used.
- The live dataset is what this website, the
API, the MCP server and the downloads below serve.
It changes with every deploy. Its version is reported as
dataset_versionin every API response. It is a plain release number (for example2.0.1) only when the live content is identical to that archived release, checked by a content fingerprint on every deploy. Otherwise it reads2.0.1+unreleased, meaning "changed since v2.0.1, not yet archived". To cite the live state exactly, give thecorpus_snapshotandsource_commitfrom the API response. - An archived version is an immutable Zenodo deposit with its own version DOI. Cite one when your analysis must be reproducible from a file that will never change; download it and analyse that file, not the live API.
Cite this resource
If you use this dataset, please cite it as:
Barton, O. (2026). Oliver's mTOR Atlas [Data set]. Zenodo. https://doi.org/10.5281/zenodo.22059963
A machine-readable citation file is also available: CITATION.cff.
Download the data
The full corpus as flat CSV/JSON files -- the same data behind every page on this site, without scraping HTML. Regenerated on every deploy, so file sizes below reflect the current corpus, not a stale snapshot.
- studies.csv (text/csv, 1145 KB)
- studies.json (application/json, 1402 KB)
- entities.csv (text/csv, 89 KB)
- entities.json (application/json, 119 KB)
- relations.csv (text/csv, 117 KB)
- relations.json (application/json, 176 KB)
These exports are living data, not a permanent citable snapshot -- for a DOI-versioned copy, use the Zenodo archive above.
What's in the corpus right now
518 hand-curated primary studies on the mTOR signaling pathway, each labelled by the kind of study behind it and linked back to its DOI or PubMed record. See About & Methodology for how a study earns a place and how the grading works.
Programmatic access
Every record on this site is also available as JSON through a read-only API (no key, CC BY 4.0, OpenAPI 3.1). The API exposes the pathway as claims: each relation carries its mechanism, the conditions it holds under, and the studies that support or contradict it. AI assistants can use the same data through an MCP server.
Thank you. Your answer was counted anonymously.