Written and maintained by CASRAI Editorial Board
Last updated
NCBI Protein is the National Center for Biotechnology Information’s database of protein sequence records, at ncbi.nlm.nih.gov/protein. Researchers and librarians use it to find a sequence by accession, protein name or organism, then follow links from the record to genes, conserved domains, structures and literature. This page explains what the database holds, which search fields it supports, and when the related Identical Protein Groups resource is the better place to look.
What NCBI Protein is
NCBI’s own database description for Protein is “Protein sequence record” (NCBI E-utilities EInfo, read 2026-11-08). The re3data registry describes it as a collection of sequences from several sources: translations from annotated coding regions in GenBank, RefSeq and TPA, plus records from SwissProt, PIR, PRF and PDB (re3data record r3d100010776, last updated 2026-07-08).
It is a sequence-record database, not an alignment tool. Protein BLAST and the 3D structure viewer are separate NCBI tools.
Who runs it and what it covers
NCBI, part of the US National Library of Medicine at the National Institutes of Health, runs the database. According to EInfo, the Protein database held 1,662,422,025 records, last updated 2026-11-07 (build 261106). The count changes daily, so treat the figure as a dated snapshot.
Per re3data, the record sources are GenBank, RefSeq and TPA translations, SwissProt, PIR, PRF and PDB. Because accessions come from several source systems, one underlying sequence can appear in more than one record.
How to search NCBI Protein
Besides free-text search, the database exposes named search fields. The table lists the tags EInfo reports for the protein database (read 2026-11-08). The third column is a plain-language reading of each field’s name.
| Tag | Field name (per EInfo) | Typical use |
|---|---|---|
| ACCN | Accession | Find a record from an accession number |
| PACC | Primary Accession | Match on the primary accession |
| PROT | Protein Name | Search by protein name |
| GENE | Gene Name | Search by encoding gene name |
| ORGN | Organism | Restrict to an organism or taxon |
| PORG | Primary Organism | Restrict to the primary organism |
| ECNO | EC/RN Number | Search by enzyme or registry number |
| SLEN | Sequence Length | Filter by length |
| MLWT | Molecular Weight | Filter by molecular weight |
| FKEY | Feature key | Find records with a given annotated feature |
| PDAT, MDAT | Publication Date, Modification Date | Filter by record dates |
| GPRJ, ASSM | BioProject, Assembly | Find proteins from a project or genome assembly |
| FILT, PROP | Filter, Properties | Limit to a record subset or property |
EInfo also lists ALL (All Fields), DIV (Division), STRN (Strain), ISOL (Isolate), AUTH, JOUR, TITL, KYWD and SQID (SeqID String).
NCBI protein sequence and accessions
For the Identical Protein Groups resource, NCBI states that a protein accession link returns the record in GenPept format, and that the nucleotide coordinates link shows the coding sequence in GenBank format. Protein-only accessions (SwissProt, PDB, patent) have no nucleotide coding sequence (NCBI IPG About page, read 2026-11-08).
Where a record leads
EInfo lists these link names for the protein database: Gene Links, Nucleotide Links, Conserved Domain Links, Protein Family Models, Protein Cluster Links, Structures and Identical Structures, PubMed Links, SNP Links, Taxonomy Links, OMIM Links, cross-links to UniProtKB and RefSeq protein records, BioAssay by Target, PubChem Compound and Substance links, Genome Links, and Encoding mRNA. From one sequence you can therefore move to its gene, its domain annotation, related 3D structures and the papers connected to it.
For the gene side of that route see NCBI Gene; for organism names see NCBI Taxonomy; for curated reference accessions see NCBI Reference Sequence; for domain and family annotation from another provider see InterPro. For predicted structures held outside NCBI, see the AlphaFold Protein Structure Database.
Protein versus Identical Protein Groups
Identical Protein Groups (IPG) is a related NCBI resource. NCBI describes it as groups of protein records in which each group represents a unique protein sequence, so result sets are smaller than in Protein (NCBI IPG About page, read 2026-11-08). IPG includes accessions from GenBank, RefSeq, SwissProt, PDB and other sources, can be searched by accession, protein name or taxonomic terms, and highlights a “best” RefSeq or SwissProt accession where one exists. Report source labels are RefSeq, INSDC, SwissProt, PDB, PIR and Patent, and a report can be downloaded as a table of accessions, coding-sequence coordinates and taxonomy.
NCBI also states that IPG searches cover all GenBank protein translations, including those from annotated WGS genomes, which will not be included in the Protein resource in the future. For proteins from a whole-genome-shotgun assembly, IPG is the resource NCBI points to.
Strengths and limits
- Strength: a very large multi-source collection, 1.66 billion records as of 2026-11-07, with open access.
- Strength: cross-links to genes, nucleotides, domains, structures, taxonomy and literature.
- Limit: records come from several source databases, so one sequence can appear under many accessions; IPG exists to group identical sequences.
- Limit: NCBI states that translations from annotated WGS genomes will not be included in the Protein resource in future.
Where to look instead, or next
| Resource | What it adds, per NCBI |
|---|---|
| Identical Protein Groups | One group per unique protein sequence; smaller result sets; covers WGS translations |
| Gene Links | The gene record connected to the protein |
| Conserved Domain Links, Protein Family Models | Domain and family annotation connected to a record |
| Structures, Identical Structures | Related 3D structures |
| Protein (UniProtKB), Protein (RefSeq) links | The matching record in UniProtKB or RefSeq |
FAQ
What is NCBI in bioinformatics?
NCBI is the National Center for Biotechnology Information, part of the US National Library of Medicine. It hosts sequence, gene, taxonomy and literature databases, including the Protein database described here.
What is the NCBI protein database?
It is NCBI’s database of protein sequence records, drawn from GenBank, RefSeq and TPA translations and from SwissProt, PIR, PRF and PDB records. Access is open.
What does an NCBI protein example look like?
Records are identified by accessions, which you can search with the ACCN field. NCBI notes that RefSeq WP_ accessions denote non-redundant bacterial proteins annotated on many genomes.
What does an NCBI protein record say about function?
A record is a sequence entry with annotation. To see the evidence NCBI connects to it, use the Gene Links, Conserved Domain Links, Protein Family Models and PubMed Links.
How do I run NCBI protein BLAST?
Protein BLAST is a separate NCBI tool that searches protein databases using a protein query. Open it from the NCBI BLAST site rather than from the Protein search box.
Can I see a 3D structure from a protein record?
Use the Structures link on a record. NCBI also provides a separate tool for viewing 3D protein structures.
Sources
- NCBI E-utilities EInfo, protein database: eutils.ncbi.nlm.nih.gov/entrez/eutils/einfo.fcgi?db=protein (read 2026-11-08)
- NCBI Identical Protein Groups, About: www.ncbi.nlm.nih.gov/ipg/docs/about/ (read 2026-11-08)
- re3data.org, NCBI Protein record: www.re3data.org/repository/r3d100010776 (record last updated 2026-07-08)
