Other ways to access predicted protein structures

There are many different ways to access protein structures predicted by AlphaFold2. The 3D-Beacons Network provides standardised macromolecular structures from a host of providers, including AlphaFold2. Furthermore, AlphaFold2 and its database have been integrated into many pieces of software used to work with predicted structures.

The 3D-Beacons Network

The 3D-Beacons Network is an open collaboration of macromolecular structure model providers.

It aims to provide standardised model coordinates and meta-information from all contributing data resources on a unified platform. It also provides unified programmatic access to experimentally-determined and predicted structure models. The 3D-Beacons Network brings together providers including: PDBe, AFDB, Protein Ensemble Database, Isoform.io, SWISS-MODEL, AlphaFill, SASBDB and more.

All data on the 3D-Beacons Network is freely available for both academic and commercial use, under the Creative Commons Attribution 4.0 (CC-BY 4.0) licence.

Software packages supporting access to AlphaFold2

AlphaFold2 and the AlphaFold Protein Structure Database (AFDB) have been integrated into specialised pieces of software, used for importing, processing and further working with predicted structures.

This means that, for many practical applications, users don’t need to run AlphaFold2 predictions manually. You don’t even need to download the structures from the AFDB. These specialist programs offer automated access to AlphaFold2 structures as standard.

For example, suppose you are an experimental structural biologist interested in solving a crystal structure by molecular replacement. We recommend using automated software for crystallography like MrBUMP from the CCP4 Suite (Agirre et al., 2023) . It will automatically analyse your protein sequence, fetch the corresponding predicted structures from AFDB, and solve the structure. No user intervention is required (Simpkin et al., 2023).