PDBselect

PDBselect provides a curated, non-redundant set of representative protein chains from the Protein Data Bank (PDB) for unbiased statistical and comparative analyses by enforcing low mutual sequence identity.

Key Features:

• Curated representative chains: A static curated list of representative protein chains sampled from the Protein Data Bank (PDB).
• Low mutual sequence identity: Selected chains are chosen to have low mutual sequence identity to minimize redundancy.
• Non-redundant dataset: Produces a dataset intended to reflect the structural diversity present in the PDB.
• PDBfilter-select: Provides the PDBfilter-select service to generate custom selections from the PDB based on specified criteria.
• Bias reduction: Designed to reduce bias in statistical analyses by preventing overrepresentation of similar sequences.

Scientific Applications:

• Comparative structural analysis: Supports comparative studies by supplying diverse, representative chains for structure-based comparisons.
• Evolutionary analysis: Supports evolutionary and phylogenetic analyses by providing non-redundant sequence samples.
• Unbiased statistics: Enables unbiased statistical analysis of sequence and structural properties by minimizing redundancy.

Methodology:

Selection enforces low mutual sequence identity among chosen protein chains drawn from the Protein Data Bank (PDB).