Document archive release-report routes and unknown procedures

README.md

@@ -112,3 +112,5 @@ the repository. Europe PMC adds papers over time, so a re-run may read a differe
 ## Licence
 
 Apache License 2.0 (see [LICENSE](LICENSE)), as for Foundation0's other repositories.
+
+Draft work for #3: [archive reporting routes](notes/archive-reports.md).

notes/archive-reports.md

@@ -0,0 +1,69 @@
+# How archives handle release reports
+
+Status: source review, claimed, checked on 28 October 2026. No messages sent.
+Refs #3. These routes support preparing the private report in #5; they do not
+authorize sending it. A person must check each case and sign the report.
+
+## PRIDE
+
+Use the [PRIDE help route](https://www.ebi.ac.uk/training/online/courses/pride-quick-tour/get-help-and-support-on-pride/)
+for a third-party query. The [owner publication instructions](https://github.com/orgs/PRIDE-Archive/discussions/14)
+describe signing in and publishing the dataset. They say to contact support if
+the data remain unavailable after two to three days; very large datasets may take
+longer to copy. That concerns an owner's publication request, not a response-time
+promise for an outside report. These sources do not establish whether PRIDE first
+contacts an author after an outside report.
+
+Proposed report contents: the accession, paper DOI and publication date, exact
+data-availability quotation, timestamped archive response, and any detected typo.
+This is our proposed evidence packet, not a published mandatory PRIDE form.
+
+## MassIVE
+
+The [public-release instructions](https://ccms-ucsd.github.io/MassIVEDocumentation/make_public/)
+describe the dataset owner's release action. The [official support FAQ](https://ccms-ucsd.github.io/MassIVEDocumentation/fair_use_policy_faq/)
+lists `ccms@proteomics.ucsd.edu` for help. The [ProteoSAFe contact page](https://massive.ucsd.edu/ProteoSAFe/contact.jsp)
+lists `ccms-web@cs.ucsd.edu` for software questions; do not confuse those with a
+published third-party dataset-release workflow.
+
+Prepare the same private evidence packet, including both repository and
+ProteomeXchange identifiers when known. The reviewed pages do not state mandatory
+fields, author-contact policy, or a turnaround time for an outside release report.
+Ask the curators to confirm the appropriate route before sending a case list.
+
+## iProX
+
+The [official contact page](https://www.iprox.cn/page/helpInformation.html) lists
+`iprox@iprox.org`. The [user manual, section 5.1](https://www.iprox.cn/page/helpEn.html)
+distinguishes public, scheduled-public and private access. It says associated data
+will be released when the paper is published, including when private access was
+selected. It does not describe an outside reporter's submission form, whether an
+author is contacted first, or a response-time commitment.
+
+Prepare the accession pair, paper citation, publication date, quotation and dated
+access check privately. Ask the curators which evidence they require. Do not infer
+that a login prompt proves a release-policy violation.
+
+## GEO
+
+The [GEO FAQ](https://www.ncbi.nlm.nih.gov/geo/info/faq.html) says records cited in
+a publication or public preprint must be released, even before their scheduled
+release date. GEO staff are obligated to release such records. The
+[update instructions](https://www.ncbi.nlm.nih.gov/geo/info/update.html) give
+`geo@ncbi.nlm.nih.gov` and request the accessions when an owner asks to change a
+release date. A third-party packet should include the accession, publication or
+preprint link, exact citation, and dated private-status observation.
+
+The reviewed instructions do not promise a response time or state whether staff
+contact the author before acting on an outside report. Time estimates for routine
+record updates are not a release-report service commitment.
+
+## Before a person sends anything
+
+Recheck the route and dataset status; these pages can change. Have a second reader
+check the paper, accession and withdrawal status using #4's procedure. Remove
+uncertain cases from the first packet. Ask a neutral question about release status
+and give curators the evidence privately. Publish aggregate counts only.
+
+The remaining check for this issue is independent review of these source readings.
+Unknown procedures remain unknown until a curator replies to an approved query.