Corrections log
Substantive corrections are logged here with the date and what changed.
Substantive corrections are logged here with the date and what changed. Pages are not silently edited. Typographical fixes are not logged; anything that changes a number, a citation, or the meaning of a claim is.
17 August 2026 — Dihexa: scope of the integrity flags understated
Page: Dihexa (PNB-0408). What changed: the record described two flagged papers and one retraction. There are four flagged papers and three retractions.
The original check was run against the single citation named in the secondary literature and its immediate companion. Re-querying PubMed across the originating laboratory's publications in the Journal of Pharmacology and Experimental Therapeutics returned four Notices of Concern printed in September 2021 (PMIDs 34551987, 34551988, 34551989, 34551990) and three retractions in April 2025 (PMIDs 40312092, 40312093, 40312094). Kawas 2011 and Kawas 2012 were absent from the record entirely; both are now typed by PubMed as Retracted Publication. McCoy 2013 remains under the unresolved 2021 notice.
The page previously closed by observing that secondary pages mentioning only the 2025 retraction understate the position of the record. It was doing the same thing, at a smaller scale. The standfirst, the section covering the flags and the reference list have been rewritten, and the method that produced the error — checking a named citation rather than the author cluster around it — is recorded in the note on checking retraction status.
17 August 2026 — Errata added to four citations
Pages: BAM15, Humanin, and the note on species, route and sample size. What changed: four references now carry the erratum attached to them. No figure or conclusion on any page changed.
The dihexa correction above prompted a sweep of every PMID the site cited at that date — 194 of them — checking PubMed publication types and correction links on every one. The site has since grown, and the sweep is re-run against the whole reference list whenever it does. It returned no retraction or expression of concern outside the dihexa cluster, and four papers carrying errata that the references did not mention. Three are resolved: the BAM15 combination paper's correction redraws a chemical structure in a figure, the Button 2013 erratum corrects a definition on page 366, and the Ioannidis 2008 erratum changes one number on page 645. None touches a figure quoted here, and each reference now says so.
The fourth is not resolved. The humanin discovery paper carries a correction at Proc Natl Acad Sci USA 2001;98(22):12854 which has no PMID of its own, and its text could not be retrieved from PubMed Central, the publisher, or the issue's corrections page. The reference now records the correction's existence and states that what it changed is not established. That is the same treatment any untraceable claim gets here, applied to our own citation.
Reporting an error
If a figure here is wrong, or a citation does not support the claim attached to it, tell us and include the source. Corrections that change a substantive claim are logged within the entry and the page carries an updated review date.