DiscoTope

DiscoTope predicts discontinuous B-cell epitopes from protein three-dimensional structures to identify antigenic surface residues for vaccine design, disease diagnostics, and immunotherapeutics.

Key Features:

• Structure-based prediction: Predicts discontinuous B-cell epitopes using protein three-dimensional structures.
• Spatial neighborhood definition: Employs a novel spatial neighborhood definition to define residue neighborhoods around potential epitope sites.
• Half-sphere exposure: Uses half-sphere exposure as a surface exposure measure to assess amino acid surface accessibility.
• Updated benchmark definitions: Incorporates improved benchmark definitions to address issues present in previous methods.
• DiscoTope-2.0 performance: DiscoTope-2.0 increases average AUC from 0.791 to 0.824 on a training dataset of 13 proteins with sufficient biological information.
• Evaluation: Demonstrates improved results in both cross-validation and independent evaluations compared to other methods.

Scientific Applications:

• Vaccine design: Identifies candidate B-cell epitopes to inform rational vaccine antigen selection.
• Disease diagnostics: Pinpoints antigenic surface residues that can be targeted for diagnostic reagent development.
• Immunotherapeutics: Supports identification of epitope targets relevant to antibody- and immune-based therapeutics.

Methodology:

Uses protein three-dimensional structures and applies a novel spatial neighborhood definition together with half-sphere exposure as a surface measure to predict discontinuous B-cell epitopes.