Architect

Architect reconstructs high-quality metabolic models from protein sequences using ensemble-based enzyme annotation and likelihood-guided gap-filling for SBML-compatible constraint-based modeling.

Key Features:

• Ensemble-Based Enzyme Annotation: Integrates predictions from multiple enzyme annotation tools to compute likelihood scores for EC numbers, improving precision and recall of enzyme function assignments.
• Automated Metabolic Network Construction: Constructs metabolic networks by mapping likelihood-scored EC numbers to reactions and incorporating the resulting reaction set into the model.
• Likelihood-Guided Gap-Filling: Executes gap-filling using the same EC likelihood scores to add reactions required for model completeness and functionality.
• SBML Output for Constraints-Based Modeling: Exports reconstructed metabolic models in SBML (Systems Biology Markup Language) compatible with constraints-based modeling frameworks for simulation of growth and metabolism.

Scientific Applications:

• Systems Biology: Enables reconstruction of organism-specific metabolic networks for systems-level analyses and modeling.
• Microbial Metabolism: Supports investigation of microbial metabolic capabilities through model-based simulation and interpretation.
• Biotechnology and Synthetic Biology: Provides SBML models to inform metabolic engineering, pathway design, and synthetic biology applications.
• Computational Biology and Bioinformatics: Supplies likelihood-scored enzyme annotations and network models for comparative analyses and computational studies.

Methodology:

Accepts protein sequence input; performs ensemble enzyme annotation by combining outputs from multiple tools to compute likelihood scores for EC numbers; constructs a metabolic network from these annotations and performs likelihood-guided gap-filling; exports the final model in SBML format.