The PDB archive

Searches and reports performed on this RCSB PDB website utilize data from the PDB archive. The PDB archive is maintained by thewwPDB at the main archive, files.wwpdb.org (data download details) and the versioned archive, files-versioned.wwpdb.org (versioning details). Since February 01 2023, the wwPDB enriches PDB entries with additional annotation and distributes the latest versions of each entry via next generation archive (NextGen) accessible atfiles-nextgen.wwpdb.org.

In addition to experimental structures, the PDB archive includes structures determined by integrative and hybrid structure determination methods (IHM). Users can access and download IHM structures and associated data atfiles.wwpdb.org/pub/pdb_ihm/.

All data are available via theHTTPS protocol. Note that the FTP protocol is no longer supported. See theannouncement.

RCSB PDB hosts the archive as part of the Registry of Open Data on Amazon Web Services (AWS).

File Access URLs

DNS names are required for programmatic access to PDB archive downloads:

• HTTPS:https://files.wwpdb.org
• RSYNC: rsync://rsync.rcsb.org

Automated download of data

The URLs in this document are useful for scripted downloads using utilities such aswget. For instance you can consider using thebatch downloads shell script.

The RCSB PDB providesrsync capabilities for efficiently maintaining full copies of the archive. To facilitate automated downloads, we offer scripts that simplify the process.

Use the following script to copy the current contents of the entire archive:rsyncPDB.sh.

Additional information on obtaining and maintaining copies of the entire PDB archive or certain portions of it is available atwwpdb.org/ftp/pdb-ftp-sites.

Archival snapshots of the PDB archive

Since 2004, the PDB archive has been preserved through yearly time-stamped snapshots. These snapshots provide well-defined datasets for research on the PDB archive and include coordinate data in multiple formats, as well as experimental data.

The archival snapshots maintain the historical directory structure of the PDB archive. Coordinate files are organized into subdirectories based on the two middle characters of the PDB ID. For example, the structure 100d is located in the directory '00'.

Each file's date and timestamp reflect the last modification time, providing a historical reference for changes within the archive.

For more details on accessing these snapshots, visit thedocumentation.

Major directories in the PDB archive

The directorypub/pdb is the entry directory for the PDB archive downloads.

Some general notes:

• Entry files are date-stamped to show the date they were released
• Entries are grouped by the middle two characters of the 4-character PDB identifier. For example, entry file pdb100d.ent can be found inpub/pdb/data/structures/divided/pdb/00/pdb100d.ent.gz
• The two letter naming convention for structure holdings is retained for the directories within/pub/pdb/data/structures/divided and/pub/pdb/data/structures/divided/obsolete but not for the directories within/pub/pdb/data/structures/all, which contain the structure holdings in undivided layout.
• PDB entries are available in PDB, mmCIF, and PDBML/XML format.
• Only UNIXcompressed files are supported for coordinates, structure factors, and restraints.

For information about large structures that cannot be represented in the legacy PDB file format seehere.

Other downloads offered by RCSB PDB

Some of the http links above are also available in a short style (e.g./download/4hhb.cif.gz). Additionally, for the short style links 2 URLs are available:

• view: The HTTP/HTTPS response headers to the client are set with: Content-Type: text/plain
• download: The HTTP/HTTPS response headers to the client are set with: Content-Type: application/octet-stream and Content-Transfer-Encoding: binary

PDB entry files

PDB entry files are available in several file formats (PDBx/mmCIF, XML, BinaryCIF and legacy PDB for some entries), compressed or uncompressed, and with an option to download a file containing only "header" information (summary data, no coordinates).

The following table contains all of the legacy PDB format URLs.Please note these are to be discontinued when the PDB transitions to extended PDB IDs.

Small molecule files

Small molecule files, including the ligands/chemical components maintained in the Chemical Component Dictionary and the Biologically Interesting Molecule Reference Dictionary (BIRD) are available in multiple formats.

Experimental data files and 3DEM maps

This table includes structure factors, NMR constraints, chemical shifts, electron density maps and map coefficient files.

Sequence data

Sequence data in FASTA format (full deposited sequence as in SEQRES records).

Please note that the FASTA download service at URL/pdb/download/downloadFastaFiles.do?structureIdList=4hhb&compressionType=uncompressedhas been discontinued. Users will need to migrate to the new endpoints below. Note that the output of the new endpoints are per entity (with chain identifiers provided in header) instead of per chain.

Sequence clusters data

Results of the weekly clustering of protein sequences in the PDB byDIAMOND at 30%, 40%, 50%, 70%, 90%, 95%, and 100% sequence identity. Note that these files use polymer entity identifiers, instead of chain identifiers to avoid redundancy. The files are plain text with one cluster per line, sorted from largest cluster to smallest.

Holdings data

PDB id holdings data in json format. For more information, see thedata API documentation.

Chemical Component Dictionary (CCD) Data

A subset of properties is provided for all components from theChemical Component Dictionary (CCD) which describes chemical properties of all molecules in the PDB archive. The atom file (cca.bcif) provides the following CIF columns:atom_id,comp_id,charge, andpdbx_stereo_config. The bond file (ccb.bcif) provides the following CIF columns:atom_id_1,atom_id_2,comp_id,molstar_protonation_variant,pdbx_aromatic_flag,pdbx_stereo_config, andvalue_order.

This data can be used by theMol* ModelServer.