firestar

firestar predicts catalytic and ligand-binding residues in protein sequences by leveraging structural templates from the FireDB database and assessing alignment reliability.

Key Features:

• Automated predictions: Performs automated prediction of catalytic and ligand-binding residues from sequence alignments.
• High-throughput analysis: Supports high-throughput analysis of large sets of protein sequences.
• Alignment methods: Uses HHsearch and PSI-BLAST to generate alignments between target sequences and structural templates.
• FireDB integration and expansion: Integrates structural templates and ligand annotations from FireDB, with expanded coverage including additional HHsearch-generated alignments.
• Alignment reliability assessment: Assesses alignment reliability to inform prediction confidence for functional residues.

Scientific Applications:

• Functional residue prediction: Identifies residues involved in catalytic activity and ligand binding in proteins.
• Protein function annotation: Supports interpretation and annotation of protein function from sequence data.
• Drug discovery: Facilitates identification of ligand-binding residues relevant to drug-design and target characterization.
• Benchmarking and evaluation: Has been evaluated in the CASP ligand-binding prediction experiment for method assessment.

Methodology:

Alignments between target sequences and FireDB templates are generated using HHsearch and PSI-BLAST; FireDB template annotations and ligand biological-relevance classifications are used along with alignment reliability assessments to predict catalytic and ligand-binding residues.