PiQSi: protein quaternary structure investigation

Area of Science:

• Structural biology
• Bioinformatics

Background:

• Determining the biological quaternary state of protein complexes is crucial for understanding their function.
• The Protein Data Bank (PDB) contains vast amounts of structural data, but accurate quaternary state information requires careful curation.

Purpose of the Study:

• To introduce PiQSi, a framework for manual investigation and annotation of protein quaternary structures in the PDB.
• To improve the accuracy of quaternary state determination and provide a resource for benchmarking prediction methods.

Main Methods:

• Development of the PiQSi framework for browsing and annotating protein quaternary structures.
• Manual annotation of over 10,000 structures from the PDB Biological Unit.
• Benchmarking the quality of manual annotations and assessing the need for curation.

Main Results:

• Over 10,000 PDB structures have been annotated for their quaternary state using PiQSi.
• Approximately 15% of the annotated structures had their quaternary state corrected, highlighting the need for manual curation.
• Annotations are of high quality, particularly for ambiguous cases like proteins in equilibrium between states.

Conclusions:

• PiQSi facilitates high-quality manual curation of protein quaternary structures in the PDB.
• The generated annotations enhance structural analyses and serve as a benchmark for prediction tools.
• A community-based curation system aims for a complete description of biological quaternary states in the PDB.