Phyre

Phyre predicts three-dimensional (3D) protein structures and supports ligand binding-site prediction and amino-acid variant analysis using remote homology detection for structural and functional inference.

Key Features:

• Homology detection: Uses remote homology detection through profile-profile matching and alignment of hidden Markov models with HHsearch to construct 3D protein models.
• Ligand binding and variant analysis: Predicts ligand binding sites and analyzes effects of amino-acid variants, including nonsynonymous single nucleotide polymorphisms (nsSNPs), on protein structure and function.
• Batch processing and automated searches: Supports submission of large numbers of sequences and performs automated weekly searches for challenging-to-model proteins.
• Computation time: Structure prediction tasks typically complete within approximately 30 minutes to 2 hours after submission.

Scientific Applications:

• Novel protein structure prediction: Predicts tertiary and secondary structures and domain composition for proteins lacking experimentally determined structures.
• Protein function inference: Infers protein function through ligand binding-site prediction and structural modeling.
• Variant impact analysis: Assesses the structural and functional consequences of genetic variants such as nsSNPs.

Methodology:

Employs profile-profile matching algorithms and alignment of hidden Markov models using HHsearch for remote homology detection to enable 3D model construction even when direct homologs are not apparent.