EPSVR

EPSVR predicts B-cell antigenic epitopes on protein surfaces using conformational information to identify discontinuous epitopes for immunologic research and medical applications.

Key Features:

• Discontinuous Epitope Prediction: Predicts discontinuous (conformational) B-cell antigenic epitopes located on protein surfaces.
• Support Vector Regression (SVR): Employs Support Vector Regression as the core machine learning algorithm for epitope score prediction.
• Conformational Information: Incorporates protein conformational data to inform prediction of non-linear epitope regions.

Scientific Applications:

• Immunologic Research: Facilitates study of immune recognition and vaccine development by identifying antigenic sites recognized by B-cells.
• Medical Applications: Supports therapeutic antibody design, drug design, and personalized medicine by locating specific antigenic epitopes.

Methodology:

EPSVR was developed alongside EPCES for discontinuous epitope prediction and was benchmarked using an independent test set comprising antigens without complex structures with antibodies; the epitopes in that dataset were identified through biochemical experiments.