Skip to content

For press & researchers

About the atlas, in one page

A short primer for journalists, researchers, and editors who need to understand what the ME/CFS Atlas is, how it is built, and what to cite it for. For the full methodology see /methodology. For the editorial contract see /editorial-policy.

What the atlas is

A structured, machine-generated index of published myalgic encephalomyelitis / chronic fatigue syndrome (ME/CFS) research. PubMed is queried daily; matching papers are classified (evidence level, research approach, PEM criterion, paradigm framing), summarised from title and abstract by a large language model, screened by a deterministic wording scanner, passed through an auto-moderation lint gate, and published at stable URLs under /evidence.

It is not a systematic review, a clinical resource, or a guideline. It is not a substitute for the primary paper. It is a reading aid that sits on top of the published record.

Who runs it

An independent civic-technology project, built and maintained by an independent operator. The operator is not a clinician, not a researcher, and not a journalist. The atlas does not represent any research group, patient organisation, clinic, or public body. There is no external funding and no commercial relationship; if that ever changes, it will be disclosed on /editorial-policy.

Transparency mechanisms

  • Versioned prompts and scanner rules. Every per-study summary is produced by a generator with an explicit version string; every draft is screened by a scanner with an explicit rule version. Current versions are shown in the footer of every page and on /methodology.
  • Published PubMed query. The exact search term used to seed the corpus is published verbatim on /methodology §4.
  • Retraction visibility. Retracted and withdrawn studies stay reachable at their URLs with a prominent red banner; the atlas’s known retractions are indexed on /retractions. Detection combines PubMed title prefixes and Crossref updated-by metadata (which includes Retraction Watch data).
  • Integrity checks. A read-only invariant checker runs over the corpus and catches cached-count drift, orphan topic states, retraction provenance gaps, and retraction-edge consistency.
  • Revision history per study. Every machine edit to a published study is logged with a reason, actor, and date.

What to trust it for — and what not to

The atlas is useful as a structured entry point into ME/CFS research: classifications, PEM-criterion status, paradigm labels, topic tags, and citable links to the primary paper. Every summary is explicitly machine-drafted and explicitly not peer-reviewed; that disclosure appears on every study page.

Do not trust per-study summaries as clinical advice. Do not treat a classifier-assigned paradigm label as a quality judgement. Do not quote a machine-generated summary as a source; quote the primary paper instead, using the Cite this study block on each evidence page.

Quick facts

  • 6,119 non-retracted public studies
  • 6,129 total accessible entries (10 flagged as retracted or withdrawn)
  • 16 research topics
  • Seeded from a single, published PubMed query (see /methodology §4)
  • Updated daily via scheduled ingestion cron; per-study page carries its ingest / publish / last-review timestamps

Data for reporting

  • /api/export/studies.json — paginated or full-dump JSON. Retracted studies included with an isRetracted flag.
  • /api/export/studies.csv — flat CSV variant for spreadsheet / R / Python.
  • /feed.xml — RSS 2.0 feed of recently added studies.
  • /methodology — canonical methodology statement, PubMed query, version stamps, limitations, and methodology changelog.

How to cite

Cite the primary paper for any factual claim. Cite the atlas only if you are specifically referring to the atlas’s reading or classification of a paper. Each evidence page carries a Cite this study block with APA and BibTeX forms. For an atlas-snapshot reference use the version-stamped line on /methodology §7.

Contact

For interview requests, fact-checking, or any substantive question about the atlas, use /contact. For a specific study error, use the report link on that study’s page. The atlas is maintained independently with limited capacity; response times can vary.