Consensus

Consensus produces consensus sequence alignments by integrating outputs from five distinct alignment methods to improve alignment reliability and identify regions suitable for comparative modeling of protein structures.

Key Features:

• Consensus integration: Integrates outputs from five distinct alignment methods to generate a single consensus alignment.
• Position-specific reliability: Computes a reliability measure (confidence score) for each position in the alignment.
• Low-identity performance: Operates effectively with target–template sequence identity as low as 5%.
• Modeling-region prediction: Predicts regions suitable for comparative modeling based on per-position reliability scores.
• Template handling: Accepts an optional template specified by PDB code and performs an automated structural template search if no template is provided.
• Validated accuracy: Validated on homologous protein structure pairs and yields models that typically achieve root-mean-square deviations within 3 Å from crystal structures.

Scientific Applications:

• Comparative (homology) modeling: Provides high-quality alignments and reliable-region predictions to support model building.
• Modeling low-identity homologs: Enables comparative modeling for target–template pairs with sequence identity down to 5%.
• Model assessment and selection: Supplies per-position confidence scores to guide selection of regions for structural modeling and to assess model reliability, contributing to models near 3 Å rms of crystal structures.

Methodology:

Combines outputs from five distinct alignment methods (multiple alignment strategies) to produce a consensus alignment and computes per-position confidence scores; accepts a target protein sequence and an optional template specified by PDB code, and performs an automated structural template search if no template is provided.