IntFOLD

IntFOLD predicts three-dimensional protein structures, intrinsic disorder, domain boundaries, ligand-binding sites and model quality directly from amino acid sequences to support structural and functional characterization.

Key Features:

• Tertiary Structure Prediction (nFOLD4): Predicts three-dimensional protein structures from amino acid sequences.
• Model Quality Assessment (ModFOLD 3.0): Provides quality scores for predicted models to assess reliability and accuracy.
• Intrinsic Disorder Prediction (DISOclust 2.0): Predicts intrinsically disordered regions that lack fixed tertiary structure.
• Domain Prediction (DomFOLD 2.0): Identifies structural and functional domains within protein sequences.
• Ligand Binding Site Prediction (FunFOLD 1.0): Predicts potential ligand-binding sites on protein structures.
• CASP-format machine-readable output: Produces raw machine-readable data files adhering to Critical Assessment of Methods for Protein Structure Prediction (CASP) standards.

Scientific Applications:

• Structural biology: Supports structural characterization and interpretation of protein folds.
• Drug discovery: Aids identification of ligand-binding sites relevant to drug design.
• Functional genomics: Helps annotate protein function and domains in genomics studies.
• Exploration of uncharacterized proteins: Enables sequence-driven prediction and analysis of unknown proteins.
• Experimental design and hypothesis generation: Provides predictions to inform experimental planning and hypothesis development.
• Benchmarking: Demonstrated competitive performance in CASP9, supporting the reliability of its predictions.

Methodology:

Uses nFOLD4 for tertiary-structure prediction, ModFOLD 3.0 for model quality assessment, DISOclust 2.0 for intrinsic disorder prediction, DomFOLD 2.0 for domain prediction and FunFOLD 1.0 for ligand-binding site prediction, and outputs machine-readable files in CASP format.