NCBI RefSeq: What a Reference Sequence Is and How to Find and Cite One

The NCBI Reference Sequence (RefSeq) collection is NCBI’s set of reference records for genomic DNA, transcripts and proteins. This guide shows how to find the RefSeq record for a gene, how to read the accession prefix (NM_, XM_, NG_ and the rest) so you know whether a record is curated or computed, and how to cite one so a reader can reproduce your work. The details follow NCBI’s own About RefSeq page and its how-to on finding a curated sequence record (checked 2026-10-29).

What RefSeq is used for in research

NCBI describes RefSeq as “a comprehensive, integrated, non-redundant, well-annotated set of sequences, including genomic DNA, transcripts, and proteins,” and says RefSeq sequences “form a foundation for medical, functional, and diversity studies.” NCBI describes RefSeq as a stable reference for genome annotation, gene identification and characterization, mutation and polymorphism analysis (especially RefSeqGene records), expression studies and comparative analyses. The collection covers archaea, bacteria, eukaryotes and viruses.

RefSeq and GenBank

NCBI states that RefSeq genomes are copies of selected assembled genomes available in GenBank, and RefSeq transcript and protein records are produced by computation, manual curation, or propagation from annotated genomes submitted to members of the International Nucleotide Sequence Database Collaboration (INSDC). The features NCBI singles out are non-redundancy, explicitly linked nucleotide and protein sequences, a distinct accession series, and ongoing curation with reviewed records indicated. For the archival side, including how GenBank and RefSeq assembly accessions differ, see our GenBank guide. To deposit raw reads rather than look up references, see submitting sequence data to NCBI SRA.

Access: where RefSeq records live

RefSeq needs no installation. NCBI says it is accessible through BLAST, Entrez and the NCBI FTP site, and that information also appears in NCBI’s Assembly, Genomes and Gene resources, with Genome Data Viewer available for some organisms. Use the official NCBI site (ncbi.nlm.nih.gov) as the starting point.

Step by step: find the RefSeq record for a gene

NCBI’s how-to offers three starting points. Menu and section names below are as they appear on that NCBI page.

Starting from a gene name or symbol

1. Search the Gene database with the gene name, product name or symbol. With a symbol and species, NCBI’s example query is tpo[sym] AND human[orgn].
2. Click the gene you want.
3. In the Table of Contents at the upper right of the gene record, click “Reference Sequences”.
4. In the NCBI Reference Sequences section, follow the links to the curated records for the genomic region, the transcripts and the proteins.

NCBI notes that transcript sequences are not produced for prokaryotes, so do not expect a transcript record for a prokaryotic gene.

Starting from an accession number

1. Run an All Databases search with the accession (NCBI’s examples are U00001 and AAA60471).
2. Click the result in the relevant database (Nucleotide, EST, GSS or Protein).
3. In the “Reference sequence information” section on the right of the record, follow the links to the corresponding reference sequences.
4. If a “More about the … gene” section is present, it leads to the gene record, which may list additional reference sequences.

Starting from a nucleotide or protein sequence

1. From the BLAST home page, choose the appropriate link in the Basic BLAST section.
2. Pick the form and database that match what you want back: a nucleotide query against “Reference mRNA” or “Reference genomic” uses nucleotide blast; a protein query against “Reference protein” uses protein blast.
3. Click the “BLAST” button and read the matching sequences.

NCBI says the translating BLAST services are not needed here, because the protein reference is reachable from the matching nucleotide record and the reverse. For primer design against these records, see NCBI Primer-BLAST.

Reading the accession prefix: curated or computed?

All RefSeq accessions contain an underscore. The prefix tells you how the record was made, and that matters when you write a methods section.

Prefix	What NCBI calls it	Origin, per NCBI
NM_, NR_, NP_	Known RefSeq	RNA and protein products mainly derived from GenBank cDNA and EST data and supported by the RefSeq eukaryotic curation group
XM_, XR_, XP_	Model RefSeq	RNA and protein products generated by the eukaryotic genome annotation pipeline; derived from genomic sequence, with varying levels of transcript or protein homology support, and not subject to further manual curation
NG_	Genomic region	Records for genomic regions provided to support more comprehensive genome-level annotation

Review status on a record

Curation status is shown in the COMMENT block of the RefSeq record. NCBI names these statuses: INFERRED, PREDICTED and PROVISIONAL for initial automatic records; VALIDATED after sequence-level review; and REVIEWED once additional annotation, a summary description or other functional information has been applied. NCBI also says curation is ongoing, that RefSeq transcript and protein records for a subset of organisms, primarily mammals, are curated by NCBI staff, and that some records have not been reviewed yet, so check the status instead of assuming a record is curated.

Citing a RefSeq record so others can reproduce it

Report the full accession, the database it came from (RefSeq, not GenBank) and the date you retrieved it. NCBI states that records are updated to reflect current knowledge and that both the nucleotide and protein record may change during curation, so an accession alone may not pin down the exact sequence you used. Because NCBI describes model records (XM_, XR_, XP_) as not subject to further manual curation, name the prefix class in your text so readers know what kind of record you used. For how identifiers differ from DOIs, see accession numbers vs DOIs.

Troubleshooting

• No transcript records for your gene: NCBI says transcript sequences are not produced for prokaryotes.
• Your accession has no “Reference sequence information” section: use the “More about the … gene” link, when present, to reach the gene record, which may list further reference sequences.
• Not sure which database to select in BLAST: match query type to the Reference mRNA, Reference genomic or Reference protein database, as in the steps above.
• A record looks different from the one in an older paper: NCBI states that RefSeq is updated as new data become available and that curation can change both nucleotide and protein records. Record the retrieval date and re-check the current record.

Other ways to reach RefSeq data

Route	What NCBI says it offers
BLAST	Similarity search against the Reference mRNA, Reference genomic and Reference protein databases
Entrez	RefSeq is accessible through Entrez; NCBI defines special properties to facilitate Entrez-based retrieval
NCBI FTP site	RefSeq releases and RefSeq Genomes are available for download
Assembly, Genomes and Gene resources	RefSeq information is available within these NCBI resources
Genome Data Viewer	Genome browser with additional information for some organisms

FAQ

What is the NCBI RefSeq database?

It is NCBI’s collection of non-redundant, annotated reference sequences for genomic DNA, transcripts and proteins, with ongoing curation by NCBI staff and collaborators.

What is the difference between RefSeq and GenBank?

GenBank is the archival nucleotide database; NCBI states that RefSeq genomes are copies of selected assembled genomes available in GenBank, with RefSeq adding non-redundancy, linked nucleotide and protein records and curation. See the GenBank guide for the archival record and assembly accession detail.

What is the difference between NM_ and XM_ accessions?

NM_ is a Known RefSeq transcript, mainly derived from GenBank cDNA and EST data and supported by the RefSeq eukaryotic curation group. XM_ is a Model RefSeq transcript generated by the eukaryotic genome annotation pipeline and not subject to further manual curation.

How do I find the reference sequence for a gene?

Search the Gene database by symbol and species, open the gene record, then click “Reference Sequences” in its Table of Contents.

Sources

• NCBI, About RefSeq: https://www.ncbi.nlm.nih.gov/refseq/about/ (checked 2026-10-29).
• NCBI, How to: Find a curated version of a sequence record (NCBI Reference Sequence): https://www.ncbi.nlm.nih.gov/guide/howto/find-curated-sequence/ (checked 2026-10-29).