Epitope Mapping

Epitope mapping is divided into two fundamental approaches. Structural mapping identifies the physical contact residues between antigen and antibody, typically through X-ray crystallography, cryo-EM, or NMR spectroscopy. Functional mapping identifies residues that, when mutated, abrogate binding, often using alanine scanning mutagenesis or deep mutational scanning. A residue can be structurally proximal but functionally silent, meaning it makes contact but does not contribute energetically to the binding interface. Computational methods must distinguish between these two categories to accurately predict the binding hot spot residues that dominate the free energy of binding.

Epitopes are classified by their structural nature. Linear epitopes consist of a contiguous stretch of amino acids in the primary sequence, typically 5-15 residues long, and are often recognized by T-cells via MHC presentation. Conformational epitopes are composed of residues that are spatially proximal in the folded 3D structure but distant in the primary sequence. Approximately 90% of B-cell epitopes are conformational, making their prediction heavily dependent on accurate antigen structure models. Computational mapping must account for side-chain orientation and backbone flexibility to correctly assemble discontinuous patches.

Modern in silico epitope mapping employs diverse algorithmic strategies:

• Structure-based docking: Uses physics-based scoring functions or deep learning to predict antibody-antigen complex structures and extract contact residues.
• Solvent accessibility analysis: Identifies surface-exposed residues likely to be available for antibody binding.
• Sequence-based predictors: Machine learning models trained on known epitope databases (e.g., IEDB) to predict antigenic propensity from sequence alone.
• Molecular dynamics simulation: Captures conformational sampling to identify transient epitopes that are not visible in static crystal structures.
• Graph neural networks: Model the antigen surface as a mesh to predict binding propensity scores per residue.

The antibody paratope and antigen epitope exhibit shape complementarity and chemical complementarity. Shape complementarity refers to the geometric fit between the protruding CDR loops (especially CDR-H3) and the antigen surface topography. Chemical complementarity involves the pairing of hydrogen bond donors/acceptors, salt bridges, and hydrophobic patches. Computational epitope mapping often employs interface analysis tools that calculate buried surface area (BSA) upon complex formation, with a typical antibody-antigen interface burying 1,500-2,300 Ų. Residues contributing >10 Ų of BSA are considered part of the structural epitope.

Epitope binning is a high-throughput experimental technique that groups antibodies into bins based on whether they compete for the same or overlapping epitopes. This is measured via surface plasmon resonance (SPR) or biolayer interferometry (BLI) in a pairwise sandwich assay. Computationally, epitope binning can be predicted by clustering docked antibody-antigen complexes based on the spatial overlap of their predicted epitopes. This is critical for selecting antibody pairs for sandwich immunoassays or bispecific antibody formats, where steric hindrance between the two arms must be avoided.

In therapeutic antibody development, epitope mapping is a cornerstone of intellectual property strategy. A novel epitope can be patented independently of the antibody sequence, providing broader protection. Computational epitope mapping enables the rapid comparison of a candidate antibody's binding footprint against all known antibodies targeting the same antigen, identifying potential freedom-to-operate conflicts. The World Intellectual Property Organization (WIPO) increasingly accepts structurally defined epitopes as patentable subject matter, making high-resolution computational mapping a critical business function beyond its scientific utility.

The specific set of amino acid residues on the antibody's variable domain that directly interacts with the epitope. It is primarily composed of the six complementarity-determining regions (CDRs) , with the heavy chain CDR3 often playing a dominant role in binding specificity. Computational paratope prediction tools, such as Parapred, use deep learning on antibody sequence and structure to identify these contact residues without requiring the antigen structure.

A critical distinction in epitope classification:

• Linear Epitopes: A continuous sequence of amino acids recognized by the antibody in its primary structure. These are often mapped using peptide scanning arrays.
• Conformational (Discontinuous) Epitopes: Residues that are scattered across the primary sequence but brought into spatial proximity by the protein's three-dimensional folding. The vast majority of antibody epitopes are conformational, requiring structural biology techniques like X-ray crystallography or cryo-EM for definitive mapping.

A powerful experimental technique for mapping epitopes in solution. The method measures the rate at which backbone amide hydrogens exchange with deuterium in a solvent. When an antibody binds an antigen, the binding interface is protected from exchange, creating a distinct 'footprint' of reduced deuteration. This differential pattern, detected by mass spectrometry, reveals the epitope location without requiring crystallization.

A systematic method to identify binding energy hot spots within an epitope. Each residue in the suspected interface is individually mutated to alanine, which truncates the side chain. A significant drop in binding affinity upon mutation indicates that the original residue is a critical contributor to the interaction. This data is often used to train machine learning models to predict the energetic consequences of mutations.

A computational simulation that predicts the three-dimensional binding pose of an antibody relative to its antigen. Methods range from physics-based molecular docking (e.g., ClusPro, HADDOCK) to deep learning-based predictors (e.g., AlphaFold-Multimer, EquiDock). Accurate docking directly reveals the epitope and paratope by identifying the interface residues in the predicted complex structure.

The computational forecasting of specific viral mutations that would allow a pathogen to evade neutralization by a given antibody. This is a functional extension of epitope mapping that uses deep mutational scanning data and language models to predict how mutations at epitope residues impact antibody binding. This analysis is crucial for assessing the durability of antiviral therapeutics and designing variant-proof vaccines.