RCSB Protein Data Bank: Efficient Searching and Simultaneous Access to One Million Computed Structure Models...

Area of Science:

• Structural Biology
• Bioinformatics
• Computational Biology

Background:

• The Research Collaboratory for Structural Bioinformatics Protein Data Bank (RCSB PDB) is a vital resource for experimentally-determined 3D biomolecular structures.
• Despite decades of growth, experimentally-determined structures cover only a fraction of known protein sequences.
• Machine learning models now predict protein structures with accuracy comparable to low-resolution experimental methods.

Purpose of the Study:

• To integrate Computed Structure Models (CSMs) with experimentally-determined structures in the RCSB PDB.
• To enhance the RCSB PDB cyberinfrastructure for scalable delivery of a vastly increased volume of structural data.
• To provide users with comprehensive access to both experimental and predicted protein structures.

Main Methods:

• Incorporation of approximately 1,000,000 CSMs from AlphaFold DB and ModelArchive.
• Integration of CSMs alongside existing experimental PDB structures (approximately 200,000).
• Updates to RCSB PDB Data, Search, and 1D-Coordinates application programming interfaces (APIs) for backward compatibility and opt-in CSM access.

Main Results:

• RCSB.org now offers access to both experimental PDB structures and a large collection of CSMs.
• API enhancements facilitate seamless integration of CSMs for programmatic users.
• The infrastructure is prepared for a sixfold increase in 3D biostructure data delivery, with capacity for hundreds of millions of CSMs.

Conclusions:

• The expanded RCSB PDB provides unprecedented access to complementary structural information across proteomes.
• Infrastructure modifications enable scalable delivery of diverse structural data, supporting a wider range of biological research.
• This integration represents a significant advancement in making comprehensive protein structural data readily available to the scientific community.