BOCTOPUS

BOCTOPUS (BOCTOPUS2) predicts transmembrane β-barrel topology and residue orientation in proteins by identifying barrel domains using dyad-repeat patterns of lipid-facing and pore-facing residues.

Key Features:

• Topology Prediction: Predicts transmembrane β-barrel topology for protein sequences.
• Barrel Domain Identification: Identifies the barrel domain boundaries within proteins.
• Residue Orientation Prediction: Predicts orientation of residues in transmembrane β-strands (lipid-facing vs pore-facing).
• Dyad-Repeat Pattern Utilization: Leverages the dyad-repeat pattern observed in lipid-facing and pore-facing residues to enhance prediction accuracy.
• Benchmark Performance: In cross-validation on a benchmark set of 42 proteins, correctly predicted topology in 69% of cases.
• Improved Accuracy Over Predecessor: Demonstrates an improvement of over 10% in topology prediction compared to the predecessor.
• False-Positive Reduction: Reduces false positives by distinguishing transmembrane β-barrels from non-transmembrane structures.

Scientific Applications:

• Membrane Protein Structural Analysis: Supports structural studies of outer membrane β-barrel proteins.
• Functional Inference and Drug Design: Informs functional analyses and structure-based drug design by providing topology and residue orientation data.
• Protein Classification: Aids classification of sequences as transmembrane β-barrels versus non-transmembrane proteins to reduce erroneous annotations.

Methodology:

Uses analysis of dyad-repeat patterns in lipid-facing and pore-facing residues of transmembrane β-barrels and was evaluated by cross-validation on a benchmark set of 42 proteins.