In addition to unique record identifiers, NCBI staff (or collaborators) often assign accession numbers to database records or individual units of data. You will encounter accession numbers mostly in databases that serve as primary repositories of sequence and other molecular data.
1. Several NCBI databases use the following format for accession numbers: [alphabetical prefix][series of digits]
For example, PRJNA318322 is an accession number for a record in the BioProject database. Other databases with similar record accessioning are: BioSample, Sequence Read Archive (SRA), GEO DataSets, dbSNP, dbVar, and Probe.
2. Accession formats for sequence records include a version number: [alphabetical prefix][series of digits][.][version number]
Version numbers are assigned for those types of sequence data that we expect to be updated over time. For example, the version suffix "3" in accession NM_000680.3 indicates that the sequence in the record has been updated twice. The version increments only if the sequence itself is updated, and does not change for updates to any other fields, such as publication lists, author names, and feature annotation on the sequence.
The formats of sequence accession numbers are of distinct types, depending on the NCBI database. The alphabetical prefixes contain various embedded information. Select a database link below for details on its accession numbers: