This page is a dated ledger of changes to the infrastructure behind peptide research documentation: the sequence databases, metadata services, article archives, and analytical guidance that a reader cites, retrieves, or revisits when documenting a research material. It is not a feed of new compound findings, and it summarizes no biological result. Each entry records what a publisher announced, when the change took effect, which records it touched, what a reader should now do differently, and what did not change. The source cut-off for every entry is September 20, 2026, Pacific time.
A funding identifier that changes format, a database that issues a new release, a download path that disappears, or a viewer that is replaced can each break a citation, a retrieval script, or a notebook screenshot without changing a single scientific fact. Knowing the date and scope of each change lets a reader repair the record instead of guessing whether the paper, sequence, or structure moved. The existing research guides on this site cover how to read a certificate and how to match a lot to its record. This page covers how to keep the external references those records point to from going stale.
What does a research-documentation ledger track, and what do its two dates mean?
This ledger tracks changes published by the operators of the databases, metadata services, archives, and guidance bodies that peptide research documentation depends on. A change qualifies for an entry when a primary source page states what changed, when it took or takes effect, and which records it touches, and when the change alters how a record is identified, retrieved, displayed, or cross-referenced. Nothing qualifies because it is scientifically interesting. Every entry uses the same eight fields, defined in the table below so that an entry can be read on its own. Two of the fields are dates, and they are not interchangeable. The publisher announcement date is the date printed on the source notice. The local checked date is the date this page's editor read that notice. A local check date is never a source publication date, and copying it into a citation as if it were one introduces an error that the ledger exists to prevent.
A research-documentation update, as a dated source ledger uses the term, is a dated change published by the operator of a database, metadata service, archive, or guidance body that alters how a record is identified, retrieved, displayed, or cross-referenced. It is not a new experimental finding. A change earns an entry only when a primary source states the change, its effective date, and the records it touches. The consequence for a reader is always about the record, not the science.
| Field | What goes in it | What it is not |
|---|---|---|
| Publisher announcement date | The date printed on the source notice by its publisher | Not the date this page was drafted or read |
| Effective date or range | The day, week, or month the change took or takes effect, as the source states it | Not an inferred exact day when the source gives only a month |
| Record or version affected | The release number, service, page, or metadata field the change touches | Not every record the publisher hosts |
| Source URL | The primary page that states the change | Not a secondary summary or a search result |
| Local checked date | The date the editor read the source, stated separately | Not a publication date, and not a promise of currency after that day |
| Substantive change | What is different in identification, retrieval, display, or cross-referencing | Not a scientific result |
| Practical relevance | What a reader documenting a research material should do differently | Not a recommendation about the material itself |
| Unchanged assumptions | What the change explicitly leaves alone | Not a guarantee that nothing else changed |
Each entry also carries a state word. A planned change has an announced future date and has not happened at the cut-off. An available change can be used now but has not replaced the old path. A completed change has replaced the old path, which may already be gone. The reader's action differs in each case: a calendar note for a planned change, a test for an available one, and a repair for a completed one.
What changed in Crossref funding metadata on September 15, 2026?
On September 15, 2026, Crossref announced a change to how funding information deposited with a scholarly record is connected to an identifier. From that date, when a depositor supplies a funder name without an identifier, Crossref matches the name to a ROR identifier. ROR, the Research Organization Registry, issues persistent identifiers for research organizations. The announcement states that previously deposited records remain unchanged and that existing Open Funder Registry identifiers, the older scheme Crossref used for funders, are not replaced by ROR identifiers. The announcement carries the DOI 10.64000/rg83r-2ft15, read directly from the page. For a reader documenting where a cited study's funding statement came from, the effect is that newly deposited records may carry a ROR identifier where earlier records of the same kind carry a bare name or an Open Funder Registry identifier. Both are valid; they differ by deposit date, not by correctness.
The reader task is preservation, not translation. A documentation record that cites a study and notes its funding should keep three things separately: the funding statement exactly as the paper prints it, the identifier the metadata carries, and the scheme that identifier belongs to. A record written before September 15 and one written after it can legitimately show different identifier types for the same funder. Rewriting the older one to match the newer one is an edit with no source behind it.
Two limits follow from the Crossref text. The announcement describes new deposits; it does not say every historical record was migrated, and this page makes no such claim. And funding metadata is provenance context. It records who supported the work as the depositor reported it. It is not evidence that any result in the paper is correct, and it is not a quality signal for any research material a documentation record describes.
What does RefSeq release 237, announced September 8, 2026, change for a citation?
On September 8, 2026, the National Center for Biotechnology Information (NCBI) announced that RefSeq release 237 was available. RefSeq is the NCBI Reference Sequence database, a curated collection of sequence records that each carry an accession number and a version suffix. A release is a dated, numbered snapshot of the whole database, distributed as files with release notes. The content snapshot the release was built from carries an earlier date, August 31, which NCBI supplies separately. The two dates answer different questions. The release number and its announcement date identify which distribution a reader downloaded. The snapshot date says when the underlying content was frozen. A reproducible citation records the release and, if a specific record was actually consulted, its accession with version and the date it was retrieved. Recording the release alone, without the accession, describes the collection and not the record.
A database release does not mean every record in the database changed. A release is a dated snapshot of the whole collection, issued on a schedule, and most records are carried forward unchanged. The NCBI notice for RefSeq release 237 announces release-specific files and notes; it does not state that any particular sequence was revised. Whether a given record changed can only be established by consulting its accession and version and comparing that version with the one cited earlier.
The version suffix is the part of an accession after the period, and it increments when the record's sequence is revised. It is the only field in the citation pattern that says whether the sequence one reader looked at differs from the sequence another reader looked at. The release number alone does not carry that information. This entry reproduces no database-wide counts or species examples, because those describe the collection rather than any record a reader would cite, and a refreshed public database establishes nothing about whether a particular peptide sequence changed or whether any supplier's item corresponds to any record.
What changed when the PMC distribution migration completed on August 26, 2026?
PubMed Central (PMC) is the NCBI archive of full-text scholarly articles. Its FTP documentation page carries a notice dated August 26, 2026, stating that the changes to its Article Dataset Distribution Service are complete and that legacy article dataset files on FTP and cloud services were removed during the week of August 24, 2026. The remaining FTP service supports identifier cross-referencing through a file named PMC-ids.csv.gz, which maps PMC identifiers to other identifiers such as DOIs. For a reader whose documentation cites a full-text article through an old bulk-download path, the consequence is that the path may now return nothing. The article itself was not touched. The archive reorganized how it distributes files, and any record that stored a file path rather than a persistent article identifier now needs a new access path recorded beside the old one, with the date of the repair.
A broken source download link does not mean a paper was retracted. Archives retire bulk-distribution paths, rename directories, and move files between services, and each breaks stored links without altering the article. The PMC notice of August 26, 2026 shows this: legacy dataset files were removed in a completed migration, and the articles remain in the archive under their identifiers. Retraction is a separate editorial action announced on the article record. Resolve the persistent identifier, record the new access path, and check the record for any update notice.
A provenance detail on this entry needs disclosure, because it is exactly the kind of thing a ledger exists to catch. The notice in the page body is dated August 26, but the page footer shows a last-modified date of August 20. Both cannot describe the same edit. This ledger uses the explicit notice date, because it is the date the publisher attached to the statement itself, and records the footer discrepancy rather than resolving it silently. A reader citing the page should do the same.
Access path and license are separate fields. The distribution change affects where a file can be fetched from and says nothing about the terms under which an article may be reused, which each article's own license sets. A completed migration also establishes nothing about the validity of any article the archive holds.
What did NCBI announce about the structure-tool redesign on August 20, 2026?
On August 20, 2026, NCBI announced redesigned versions of two structure tools: the Structure Summary page, which presents one three-dimensional macromolecular structure record, and VAST, the Vector Alignment Search Tool, which compares structures to each other. The announcement states that the updated versions are available to try, that they will fully replace the current versions in October, and that VAST+, a related comparison tool, will be retired as part of that switch. The source names the month and not a day. As of this ledger's September 20, 2026 cut-off, the October replacement is a planned change and is described that way here. A reader can look at the new pages today, but any screenshot, navigation note, or exported view from the current pages remains valid until the replacement occurs, and the date on it is what tells a later reader which version it shows.
Interface changes are the most common reason a documentation record becomes hard to follow without becoming wrong. A notebook entry that says to open a record and click a particular tab is accurate on the day it is written and may be unreproducible a month later, even though the record it points to is identical. The fix is not to avoid screenshots. It is to attach a date, and where possible a version label, to every screenshot and navigation instruction, so a later reader can tell the pre-October interface from the post-October one.
A structure viewer is research infrastructure that displays public records. A changed viewer is not a new analytical result, a changed structure record, or information about any research material or lot.
Why is the 2025 analytical-method validation guidance a reference, not a change?
This entry is deliberately older than the four 2026 announcements above, and it is included as a background reference rather than a change notice. Eurachem, a network of European analytical chemistry organizations, publishes a guide titled The Fitness for Purpose of Analytical Methods. Its third edition is labeled 2025. The guide's landing page states that the third edition covers method validation and verification, sampling and sample handling, performance characteristics, reporting, and quality control. The landing page shows an update date of March 15, 2026, which is the date the web page last changed, not the edition year of the publication. A ledger that confused the two would rewrite a 2025 document as a 2026 one. This page cites the guide as the 2025 third edition and records the page update date separately, in the same way it separates announcement dates from local check dates everywhere else.
A documentation ledger carries a guidance entry for vocabulary. When a reader records what a laboratory report claims to have measured, the words validation, verification, performance characteristic, and reporting each have a meaning in the guide's framework, and loose use makes one record harder to compare with another. The scope supports one habit in particular: define the analytical question before assessing whether any evidence answers it.
Three limits apply. The landing page was read and the full guide was not, so no numerical threshold, acceptance criterion, or procedure is asserted here. A guidance document creates no statutory obligation and certifies no supplier. And a validation framework's existence does not mean any particular laboratory's method was validated against it; only a report could say that, and the guide to reading a certificate of analysis owns how to read one.
How do you turn a source change into a useful record update?
A ledger entry is only useful if it changes something in a reader's own records. The change-impact worksheet below is a blank template for doing that. It is filled in once per affected citation or workflow, not once per announcement, because one announcement can touch several records and one record can be touched by several announcements. Its seven columns force the reader to name the old reference exactly as stored, the new reference as it should now be stored, the kind of change connecting them, which citation or workflow is actually affected, what check is required before the new reference is trusted, which claim in the record is explicitly unaffected, and the status of the repair. The unaffected-claim column is the one most often left blank in practice, and it is the one that prevents a retrieval repair from being misread later as a substantive revision of what the record says.
| Old reference | New reference | Change type | Affected citation or workflow | Check required | Unaffected claim | Status |
|---|---|---|---|---|---|---|
| The reference exactly as currently stored | The reference as it should now be stored | Access path, identifier scheme, release or version, interface, or guidance edition | The notebook entry, bibliography line, script, or procedure that uses the old reference | What must be confirmed before the new reference replaces the old one | The statement in the record that this change does not touch | Open, checked, or repaired |
Here is one worked row, hypothetical throughout. Suppose a documentation record written in 2025 supports a statement about a research material's published sequence by pointing to a full-text article, and the pointer it stored is a bulk-download path on the PMC FTP service rather than the article's persistent identifier. After the August 2026 distribution change, that path returns nothing.
| Column | Hypothetical entry |
|---|---|
| Old reference | A stored FTP file path under the legacy PMC article dataset directory, recorded in 2025 as <legacy-ftp-path> |
| New reference | The article's persistent identifier, recorded as <article-identifier>, plus the current access path as <current-access-path> |
| Change type | Access path (distribution service migration completed the week of August 24, 2026) |
| Affected citation or workflow | Bibliography line 12 of the notebook entry, and the retrieval step of the literature-check procedure |
| Check required | Resolve the persistent identifier, confirm the resolved article matches the title and authors stored in 2025, and read the article record for any update or correction notice |
| Unaffected claim | The sequence statement the article was cited for; the article's content is not stated to have changed |
| Status | Checked; repaired once the new access path is recorded beside the old one with the date of the repair |
Notice what the worked row refuses to do. It does not delete the old path, because the old path shows what was actually consulted in 2025. It does not update the sequence statement, because nothing in the distribution change touched the article. And the check-required column is entirely about confirming that the new pointer resolves to the same article, not about re-examining the science. The relevance of an infrastructure change is infrastructural, and the worksheet keeps it there.
How do you check corrections and versions before reusing a study record?
Before a study record is reused in new documentation, the reader needs to know whether the publisher has issued an update since the record was written. Crossref operates a service called Crossmark for this purpose: for participating scholarly content, it supplies a way to check publisher-reported updates such as corrections or retractions. Participating is the operative word. Crossmark reports what a participating publisher has registered, and the absence of an indicator tells a reader only that the publisher may not participate. The verification sequence below therefore uses the publisher's own article record as the primary check, available version and update information as the second, and the cited passage as the third. It applies to scholarly records only. It does not apply to laboratory certificates of analysis, which are supplier documents outside any scholarly update service; this site's available lot documentation and its certificate-reading guide handle those separately.
Absence of a Crossmark indicator is not proof that an article has no correction. Crossmark reports publisher-registered updates for participating content only, so a non-participating publisher's article can have been corrected or retracted without any indicator appearing. The dependable check is the article's own record at the publisher, read for update notices, then any version or update metadata, then a re-read of the passage the record cited. A study record should be reused only after those three checks are dated and recorded.
- Open the article record at the publisher, not a cached copy, and read it for any notice of correction, erratum, expression of concern, or retraction. Record the date this was done.
- Where the article carries version or update metadata, through Crossmark or a publisher version history, record the current version and any listed updates with their dates.
- Re-read the specific passage the documentation record cited and confirm that it still says what the record says it says. If a correction touched that passage, the record needs a dated revision, not a silent edit.
- Record the result in the change-impact worksheet from the previous section, with the unaffected-claim column filled in even when nothing changed.
This sequence is about the integrity of a citation: whether the record a reader points to still exists in the form they cited. It is not a method for judging what a study found, and no step involves interpreting results. A correction notice says the publisher changed something. What that means for any conclusion is a separate question this page does not take up.
When should a documentation ledger's date change, and when should it not?
The editorial rule for this page has three parts. An entry is added or altered only when a primary source that the editor has read supports a substantive correction, a new record, or a changed access detail; a secondary summary, a search result, or a recollection is not enough. Prior dated entries are kept rather than overwritten, so that a reader can see what was known at each cut-off. And each entry carries one of three state words, planned, available, or completed, which changes only when the source's stated date has passed and a fresh read confirms the change occurred. The October structure-tool transition is the live example: it moves from planned to completed only after a new read of the NCBI source in or after October, not because October arrives on the calendar. The rule exists because a ledger whose dates move for cosmetic reasons stops being usable as a record of what was known when.
An article's updated date should change only when a primary source, actually read, supports a substantive change to the page: a new ledger entry, a correction to an existing entry, or a changed access detail. Re-reading the same sources and finding nothing new is a check, recorded as a local checked date, not a page update. Rephrasing, reformatting, or re-rendering the page on a schedule changes nothing it says and justifies no new date. Prior dated entries stay in place when a new one is added.
| Event | Changes the page date? | What is recorded instead |
|---|---|---|
| A new primary-source announcement is read and added as an entry | Yes | The new entry with its own announcement and checked dates |
| A previously read source is re-read and found unchanged | No | A new local checked date on the affected entry |
| A planned change's stated month arrives on the calendar | No | Nothing until a fresh read confirms the change occurred |
| A fresh read confirms a planned change is complete | Yes | The entry's state word and effective date |
| The page is re-rendered, reformatted, or reworded without new facts | No | Nothing; the content is the same |
| An error in an existing entry is corrected against the source | Yes | The correction, with the prior wording preserved |
This page is drafted locally on a two-day cadence. That cadence is a work schedule, not an editorial event. It never requires a date change or a publication, and it produces no claim that the page is more current, better ranked, or more likely to be cited than any other page. Search estimates for terms such as peptides news are separate from observed visits, rankings, or citations, none of which is claimed here. The Article structured data uses the editorial date, not the date of the newest source cited.
The compact ledger, sorted by announcement date
The table below collects every entry on this page in one place, sorted by publisher announcement date with the newest first, so that a reader who wants only the dates, states, and actions can read it without the sections above. Four rows are 2026 change notices. The two rows after them are reference entries rather than change notices: the Eurachem guide is listed by its 2025 edition year with the landing-page update date beside it, and the Crossmark service is listed with no publication date because its page does not establish one. The local checked date for every row is September 20, 2026, Pacific, and the cut-off for the whole table is the same day. A reader who repeats the checks after that date should expect at least one state word to have changed, because the October structure-tool replacement was planned and not complete when this table was built.
| Announcement date | Publisher | Change | Effective date or range | State at cut-off | Reader action |
|---|---|---|---|---|---|
| September 15, 2026 | Crossref | Funder names lacking identifiers matched to ROR identifiers on new deposits | From September 15, 2026 | Completed for new deposits | Keep funding statement, identifier, and scheme as three separate fields |
| September 8, 2026 | NCBI | RefSeq release 237 available | Release September 8, 2026; content snapshot August 31, 2026 | Completed | Cite release number, accession with version if consulted, and retrieval date |
| August 26, 2026 | NCBI / PMC | Article Dataset Distribution Service change complete; legacy files removed | Week of August 24, 2026 | Completed | Replace stored file paths with persistent identifiers plus new access path |
| August 20, 2026 | NCBI | Redesigned Structure Summary and VAST available; VAST+ retirement tied to switch | Full replacement planned for October 2026, day not stated | Available to try; replacement planned | Date every screenshot and navigation note |
| 2025 edition (page updated March 15, 2026) | Eurachem | Fitness for Purpose of Analytical Methods, third edition | Current edition | Reference edition, not a change notice | Use its vocabulary when stating what a method claims to measure |
| Date not established (checked September 20, 2026) | Crossref | Crossmark update-checking service for participating content | Ongoing service | Reference service, not a change notice | Check the publisher record first; absence of an indicator proves nothing |
What the table leaves out is deliberate. There is no entry for any compound, study result, supplier, or price, because none of those is a documentation-infrastructure change. There is no completeness claim, because six sources were selected and read within a bounded window and other changes occurred elsewhere in the same period. And there is no entry dated after September 20, 2026, because nothing after that date was read.
Frequently asked questions
Is this a complete feed of peptide research news?
No. The Peptide Research Documentation Updates page is a selected ledger of changes to the databases, metadata services, archives, and guidance documents that peptide research documentation depends on. Six primary sources were read within a bounded window ending September 20, 2026, Pacific time, and only changes stated on those pages are recorded. It does not cover new compound findings, biological results, or industry announcements, and it makes no claim to have captured every infrastructure change in the period. Anything absent was outside the source set or the cut-off, not shown to be unchanged.
Which date should I cite for a database or infrastructure change?
Cite the date the publisher printed on the notice announcing the change, and record the effective date separately when the source gives one. For RefSeq release 237 that means the September 8, 2026 announcement date and, where relevant, the August 31, 2026 content snapshot date that NCBI supplies. For the PMC distribution change, cite the August 26, 2026 notice date from the page body and note that the page footer disagrees. Never cite the date you read the source as if it were the announcement date; record it as a checked date on its own line.
Does a funding-identifier update change the scientific findings?
No. The Crossref change of September 15, 2026 alters how a funder name deposited without an identifier is matched to a ROR identifier in the metadata attached to a record. It touches provenance context, meaning who is recorded as having supported the work. It does not touch the article's text, data, or conclusions, and it does not apply retroactively to records deposited earlier, which keep the funding metadata they had. A documentation record that cites a study for a finding needs no revision because of this change, at most a note that newer records may carry a different identifier scheme.
Has the announced October VAST transition already occurred at this cutoff?
No. The NCBI announcement of August 20, 2026 states that the redesigned Structure Summary and VAST pages are available to try and will fully replace the current versions in October, with VAST+ retired as part of that switch. The source names the month and not a day, and October is after the ledger's September 20, 2026 cut-off. The transition is therefore recorded as planned, the new pages as available to try, and the old pages as still current. The state will be updated only after a fresh read of the source confirms that the replacement took place.
Does loss of a legacy PMC download path mean a paper was withdrawn?
No. The PMC notice of August 26, 2026 describes a completed change to the Article Dataset Distribution Service in which legacy dataset files were removed from FTP and cloud services during the week of August 24, 2026. That removal changes where files can be fetched from and nothing else. The articles remain in the archive under their persistent identifiers, and withdrawal or retraction is a separate editorial action that appears on the article record itself. A dead legacy path is a signal to resolve the identifier, record the new access path, and check the article record for any update notice.
Does the two-day drafting schedule make this page more current?
No. The two-day cadence is a local work schedule for drafting and checking, not an editorial event. A drafting pass that re-reads the sources and finds nothing new records a checked date and changes nothing else. The ledger's stated date and its Article structured data change only when a read primary source supports a new entry, a correction, or a changed access detail. The cadence produces no publication, no claim that the ledger is more current than another page, and no expectation of ranking or citation advantage. Currency is fixed by the stated source cut-off of September 20, 2026, Pacific time.
Sources and scope notes
Every entry on this page was built from a direct read of the primary source page listed beside it, on September 20, 2026, Pacific time. That reading date is a local checked date, distinct from each source's own announcement, effective, or edition date. The Eurachem entry rests on the guide's official landing page; the full guide was not reviewed, and no numerical method assertion is drawn from it. The Crossmark entry rests on the service description page, whose own publication date is not established.
This is a selected documentation-infrastructure ledger, not a complete news feed, a compound digest, or a statement that the site is current across all topics. Dated search estimates for related query terms are separate from any observed visit, ranking, or citation, and none is claimed. No live Glow lot, laboratory, price, stock, or accreditation was verified, and the worked example rows are hypothetical teaching entries with placeholder values.
Research-only scope: this page describes how to keep external references in laboratory research documentation reproducible when the infrastructure behind them changes. It does not describe any research material, lot, or report, and it does not interpret any study's findings.
- Crossref, Improving how funding connects to research outputs, announcement dated September 15, 2026 (DOI 10.64000/rg83r-2ft15) Read September 20, 2026, Pacific.
- NCBI Insights, RefSeq Release 237 is now available, September 8, 2026 Read September 20, 2026, Pacific. Database-wide figures were not reproduced.
- PMC, FTP service page carrying the August 26, 2026 Article Dataset Distribution Service notice Read September 20, 2026, Pacific. The page footer shows a last-modified date of August 20, 2026, which disagrees with the notice date in the body.
- NCBI Insights, Exploring 3D protein structures at NCBI, August 20, 2026 Read September 20, 2026, Pacific. The source states October for the full replacement and gives no exact day.
- Eurachem, The Fitness for Purpose of Analytical Methods, third edition (2025), official guide landing page Landing page read September 20, 2026, Pacific; page update date March 15, 2026. The full guide was not reviewed.
- Crossref, Crossmark service description Read September 20, 2026, Pacific. Publication date of the page not established.