BriX

BriX provides a database of recurrent small protein fragments for protein structure analysis and modeling, enabling reconstruction of backbone segments and loop regions for applications such as backbone movement modeling and docking.

Key Features:

• Extensive Fragment Collection: The BriX database comprises over 2 million fragments of 4–14 amino acids extracted from more than 7,000 non-homologous proteins in the Astral collection and curated for quality, clustered by backbone structural similarity.
• Loop BriX Database: A specialized repository of non-regular loop elements clustered according to the end-to-end distance between the flanking regular residues to represent loop conformations.
• Clustering and Retrieval: Fragments are clustered based on backbone structural similarity to enable retrieval of structurally similar segments for modeling and analysis.
• Structural Coverage Service: A computational service that maps protein structures to BriX classes representing potential structural variations across the protein.
• Fragment Bridging Tool: A computational service that fills gaps or low-confidence regions by matching and inserting database fragments into target structures.

Scientific Applications:

• Protein structure prediction: Enhances prediction by filling structural voids with experimentally derived fragment conformations.
• Drug design: Supports analysis of protein backbones and loop regions to inform ligand-binding and design hypotheses.
• Protein dynamics and function: Enables more accurate modeling of backbone movement and loop conformational variability to study function and dynamics.

Methodology:

Proteins are segmented into small recurrent fragments (4–14 amino acids) and clustered by backbone structural similarity, with Loop BriX clustering loops by end-to-end distance between flanking regular residues; fragments are derived from and curated from the Astral protein structure collection to support retrieval for modeling tasks.