BriX: a database of protein building blocks for structural analysis, modeling and design

Area of Science:

• Structural Biology
• Bioinformatics
• Computational Chemistry

Background:

• High-resolution protein structures are crucial for understanding cellular function and drug design.
• Limited availability of protein structures hinders complex modeling tasks like backbone movement analysis and molecular docking.
• Fragment-based approaches offer an alternative strategy to address structural data limitations.

Purpose of the Study:

• To introduce the BriX and Loop BriX databases, providing extensive collections of protein fragments.
• To facilitate fragment-based strategies for protein structure analysis and modeling.
• To offer user-friendly web interfaces, APIs, and online services for accessing and utilizing these fragment resources.

Main Methods:

• The BriX database was constructed using over 7000 non-homologous proteins from the Astral collection, segmenting fragments by length (4-14 residues) and clustering by structural similarity.
• The Loop BriX database was created to classify non-regular structure elements (loops) based on the end-to-end distance of flanking regular residues.
• Both databases were made accessible online with user-friendly interfaces, APIs for high-throughput queries, and full download options.

Main Results:

• The BriX database contains approximately 2 million fragments per length, clustered by structural similarity.
• The Loop BriX database addresses the lack of loop classification in BriX, providing a resource for non-regular structures.
• Online services enable on-the-fly classification of user-submitted structures and bridging of gaps in low-confidence regions using matching fragments.

Conclusions:

• BriX and Loop BriX provide a vast resource of protein fragments to support fragment-based structural modeling and drug design strategies.
• The databases and associated online tools enhance the ability to analyze protein structure, predict variations, and complete incomplete structural models.
• These resources democratize access to fragment data, supporting a wide range of structural biology and computational chemistry applications.