What sequences does UniProtKB contain?

UniProtKB contains the sequences of proteins from reference proteomes, as well as additional selected proteins with experimental or biologically important data. Sequences from non-reference proteomes are not part of UniProtKB but are stored in UniParc and are also available via the UniProt Proteomes portal.

Where do UniProtKB sequences come from?

Most of the protein sequences provided by UniProtKB come from translations of coding sequences (CDS) submitted to the ENA/GenBank/DDBJ nucleotide sequence resources of the International Nucleotide Sequence Database Collaboration (INSDC). These CDS are either generated by gene prediction programs or are experimentally proven. The translated CDS sequences are automatically transferred to the TrEMBL section of UniProtKB. The UniProtKB/TrEMBL records may eventually be selected for manual annotation and then integrated into the UniProtKB/Swiss-Prot section.

In addition to translated coding regions from INSDC, UniProtKB protein sequences may come from:

• predicted coding regions imported from genome annotation resources such as Ensembl and RefSeq
• the PDB database of protein structures
• submissions of directly sequenced proteins
• the scientific literature

Importing and combining sequences from a range of sources means that UniProt provides a complete collection of protein sequences and contributes to consistency of protein sets across various sequence resources (Figure 3).