Project:Sourcing guidelines

How to cite trials, labels, pharmacopoeias, registries and test reports.

This guideline supplements Project:Verifiability with the specific citation practice expected in each of the source classes that recur on PeptidePedia Wiki.

Cite the publication, not the press release, and not the registry entry. Give authors (up to three, then et al.), title, journal in italics, volume(issue):pages, year, and DOI and PMID where genuinely known. Do not manufacture an identifier: an incomplete citation is recoverable, a wrong DOI sends the reader somewhere else.

Where results were announced before publication, that is itself citable — attributed as an announcement, dated, and replaced with the publication when it appears.

Cite the issuing body, the document type, the product, and the revision date. Labelling changes, so an undated label citation decays. Where a statement depends on the current label, say so and date it.

Cite by compendium, chapter number and title, noting "current revision" rather than a specific one unless the argument depends on a particular revision. Chapters are paywalled and that is not a reason to avoid citing them.

Facts about an organisation's incorporation, registered address and business scope should come from a national registry record where one is accessible. Where the record is not accessible to the editor, the fact should be attributed to whatever secondary source carries it, and flagged as unverified against a primary record.

A test report is cited as: issuing laboratory, report or sample identifier if present, analyte, date, and the route by which the report reached the wiki. The last element matters more than editors expect — a report supplied by a buyer who submitted a blind sample and a report supplied by the seller are different kinds of evidence, and the citation should make clear which it is. See Blind sampling and Chain of custody.