RRIDs and Identifying Research Resources: What the Identifier Covers and What It Doesn't

A Research Resource Identifier, or RRID, is a persistent identification number for a research resource — an antibody, a cell line, a model organism, a software tool or database, a plasmid, an instrument or core facility — that resolves to one curated record describing that resource ^[3]. Authors cite it in the methods section as "RRID:" followed by a tag naming the source authority and an accession: AB for the Antibody Registry, CVCL for Cellosaurus, MGI for Mouse Genome Informatics, SCR for the SciCrunch Registry of tools ^[4]. Its purpose is narrow and precise: to make it unambiguous which resource a paper used.

The scheme does not cover most chemical reagents, and synthetic peptides bought from a supplier generally sit outside it. This article explains what RRIDs solve, which resources carry them, how to find or request one, and what identification looks like for the compounds that fall outside. Writing the full materials statement for a purchased compound is covered in this cluster's pillar.

The problem RRIDs were built to solve

The trigger was a measurement. Vasilevsky and colleagues examined how well the resources used in published biomedical papers could be identified, and found that 54% were not uniquely identifiable, regardless of discipline, journal impact factor or the journal's reporting requirements ^[2]. Catalogue numbers change when companies merge; clone names are reused; cell lines are renamed. The paper survives, but the thing it describes can no longer be found.

The Resource Identification Initiative was the response. Its pilot engaged more than 25 biomedical journal editors, along with scientists and funding officials, and asked authors to include RRIDs in manuscripts before publication for three resource types — antibodies, model organisms, and tools such as software and databases ^[1]. The scheme has since broadened to cell lines, plasmids, instruments and core facilities ^[3].

There is evidence it does more than tidy citations. Babic and colleagues found that papers identifying their cell lines by RRID reported problematic — misidentified or contaminated — lines less often than papers that did not ^[5]. Looking a resource up in a curated registry puts any warning attached to it in front of the author.

Which resource types have identifiers

The covered types share one property: each is a discrete, distributable resource whose identity can drift or be confused — a clone, a line, a strain, a version of a program. The registries exist because a name alone was shown not to be enough ^[2][4].

What an RRID does not tell you

An RRID identifies a resource, not a batch of it. An antibody's record names the product; it does not name the vial your laboratory opened, and differences between production lots sit below the level the identifier resolves. A software tool's record identifies the project, not the version you ran. That is why the Key Resources Table keeps catalogue numbers and other identifiers alongside the RRID rather than replacing them, and why a methods section still needs the lot and the version wherever they matter ^[7].

• The lot or production run of the reagent actually used.
• The software version and the parameters it was run with.
• Whether the resource was authenticated or validated in your hands — a cell line's RRID says what the line should be, not what your culture is ^[5].
• How the material was stored, handled or prepared before use.

Read that list against a compound with no RRID and the gap between the two narrows: both need the lot, and both need the handling. The identifier fixes one layer of the problem — the product-level layer that was measurably failing ^[2]. Resolving an RRID in someone else's paper is still the fastest check a reader can make, because the curated record carries the source, the catalogue details and, for cell lines, any problems registered against the line ^[4][5].

How to find or request an RRID

1. Search the central portal, rrid.site, which indexes all currently supported resource types ^[3].
2. Use the record's own citation text. Each record carries a ready-made citation string for the methods section; copy it rather than retyping the accession ^[3].
3. If a resource is missing, use the portal's route for adding a new resource; for biological materials, deposit them with an established repository such as Addgene or ATCC, whose record then carries the identifier ^[3].
4. Cite the RRID at the first mention of the resource in the methods, in the form the journal specifies. Journals using a Key Resources Table place it in the identifier column beside the catalogue number, separated by a semicolon ^[7].

Journals that ask for them

Uptake is policy-driven. Nature Portfolio journals state support for the Resource Identification Initiative and its aim of unique, persistent identification of antibodies, cell lines, model organisms and tools ^[6]. Elsevier's Key Resources Table guidance, used across Cell Press journals, highly recommends RRIDs as the identifier wherever one exists ^[7]. Beyond that, individual journal instructions vary — some request RRIDs, some expect them for particular resource types — so the specific journal's author guidelines at submission are the authority, not a general rule.

Why most chemical reagents fall outside the scheme

The supported resource types are biological materials and tools ^[3]. Small molecules and synthetic peptides are not among them, and there is a reason: a chemically defined compound already has a complete identifier in its structure. Where an antibody's identity depends on which clone and which production run, a peptide's identity is its sequence and modifications, which can be written down exactly and confirmed analytically.

That does not solve the compound problem; it moves it. A structure identifies the molecule, not the material. Two lots of the same sequence can differ in purity, counter-ion and water content, and no structural identifier says anything about any of them. For a purchased compound, identification therefore has two layers: what the molecule is, and which physical batch was used.

Identifying a compound properly without an RRID

For the molecule, use identifiers that resolve independently of any supplier. For small molecules these are typically a CAS Registry Number and a PubChem compound identifier, which tie a name to a structure record ^[8]; the Key Resources Table's identifier column explicitly accepts CAS numbers alongside catalogue numbers ^[7]. For a peptide, the strongest identifier is the sequence itself, written with standard notation for residues, termini and modifications, together with the salt form.

For the material, use the supplier, the catalogue number and — the part journals rarely demand and reviewers increasingly ask for — the lot. A batch number's job is to single out one production run, which is exactly the layer a structural identifier cannot reach. A complete identification line for a synthetic peptide carries:

• The name and full sequence, with terminal groups and any modifications.
• The salt form, where it is known.
• A database identifier for the molecule where one exists ^[8].
• Supplier, city and country, catalogue number and lot ^[7].
• Purity with its method, and the source of that analysis.

Written that way, a compound with no RRID is identified more completely than many resources that have one: the reader knows exactly what the molecule was and exactly which batch of it was used. That is the standard the Vasilevsky study found most papers missing ^[2], and it is within reach of any author who recorded the lot on the day the material arrived.