LOCK

LOCK performs vector-based superposition of protein structures to identify distant structural similarities and conserved motifs across fold families.

Key Features:

• Hierarchical Structural Representation: Represents alpha-helices and beta-strands as vectors and compares structures across multiple tiers from secondary structure to atomic detail.
• Scoring Functions: Utilizes seven distinct scoring functions to evaluate pairs of vectors derived from different proteins.
• Dynamic Programming for Alignment: Applies a dynamic programming algorithm to identify optimal local alignments of vector sets.
• Atomic-Level Refinement: Refines vector-based alignments using atomic coordinates and iterative minimization of Root Mean Square Deviation (RMSD) between nearest atom pairs.
• Core Alignment and RMSD Minimization: Identifies a core set of well-aligned atoms and minimizes RMSD within this core to enhance alignment accuracy.
• Statistical Significance: Provides statistical significance values for alignments to assess biological relevance.

Scientific Applications:

• Comparative Protein Structure Analysis: Detects structural similarities between proteins that are not apparent from sequence comparison alone.
• Detection of Conserved Motifs and Fold Families: Identifies conserved structural motifs and fold-family relationships among distantly related proteins.
• Comparative Evaluation: Demonstrated equivalent capability to the DALI algorithm (Holm and Sander) in detecting structural similarities.

Methodology:

Initial vector alignment of alpha-helices and beta-strands using seven scoring functions and dynamic programming for local alignment, followed by atomic refinement using atomic coordinates with iterative minimization of RMSD between nearest atom pairs and RMSD minimization within an identified core set of well-aligned atoms.