GeneSilico

GeneSilico predicts protein tertiary structures from amino acid sequences or multiple sequence alignments by integrating fold-recognition (FR) methods and converting FR results into full-atom 3D models to identify remote similarities between proteins.

Key Features:

• Input formats: Accepts amino acid sequences and user-supplied multiple sequence alignments (MSAs) for prediction.
• Fold-recognition integration: Integrates multiple fold-recognition (FR) methods to detect remote homologs.
• Consensus modeling: Combines results from different FR methods and infers a consensus to enhance prediction reliability.
• Full-atom model construction: Converts fold-recognition results into full-atom 3D protein models.
• Model quality assessment: Performs uniform quality assessment of generated 3D models.
• Multi-level prediction: Produces predictions at primary, secondary, and tertiary structure levels.

Scientific Applications:

• Remote similarity detection: Identifying remote similarities between proteins via fold-recognition methods.
• Structure generation: Producing full-atom 3D models for structural analysis.
• Consensus-based prediction: Improving the reliability of structural predictions through consensus of multiple FR methods.
• MSA-enhanced prediction: Enhancing prediction accuracy by using user-defined multiple sequence alignments instead of single-sequence methods.

Methodology:

Integrates multiple fold-recognition methods, accepts amino acid sequences or user-supplied MSAs, combines FR results into a consensus, converts FR results into full-atom 3D models, and performs uniform quality assessment while producing primary, secondary and tertiary structure predictions.