AntibodyAtlas
Where does an antibody grab its target? Pick a complex below (or enter your own), and this tool fetches the structure — an experimental antibody–antigen complex from the RCSB PDB for a PDB id — then maps the interface: the epitope (the antigen residues the antibody touches), the paratope (the antibody residues that do the touching), and the CDR-like contact loops. Everything runs in your browser: the antibody and antigen chains are auto-detected from the geometry and the contacts are computed live from the coordinates; nothing is installed and nothing about you is collected.
This is the tool — the core interface geometry + honest contact mapping. A native version, when it ships, adds CDR numbering (IMGT/Kabat), a curated therapeutic-antibody library, and offline use; it never gates the web.
Paratope contact loops
Each CDR-like contact loop on the antibody, by the number of antigen contacts it makes (tallest first). In an antibody these loops are the complementarity-determining regions (CDRs); here they are found by contact geometry, so they are labelled by chain and residue range rather than by CDR number. Colour marks the antibody chain.
Interface
Structure
Methods & limits
- This is a geometric contact map, not an experimentally-mapped epitope. The epitope and paratope are the residues whose heavy atoms fall within a fixed contact distance across the antibody–antigen interface. It is not an experimental epitope (HDX, alanine scanning, escape mutagenesis) and says nothing about affinity or neutralisation.
- Chain roles are auto-detected from the geometry. The antibody is taken to be the mutually-packing chain pair (heavy + light); the antigen is the chain that pair jointly contacts. For an unusual complex the roles may be wrong, or unassignable — in that case the interface is shown generically (side vs side) rather than mislabelled.
- CDR loops are approximate. The contact loops are contiguous runs of contacting paratope residues. In an antibody these correspond to the CDRs, but they are found by geometry, not by IMGT/Kabat numbering (no ANARCI), so they are labelled by chain + residue range, not CDR-H1/H2/H3.
- Protein contacts only, single model. Waters, ions, glycans, and small-molecule ligands (HETATM) are ignored, so a glycan- or peptide-only epitope is not analysed; only the first model is read.
- Predictions have no complex. An AlphaFold model (from a UniProt accession) is a single chain, so there is no antibody–antigen interface to map — use a PDB id of an experimental complex.
About this tool
Structures come from the RCSB Protein Data Bank (and, for single-chain input, the AlphaFold Database). They are fetched directly by your browser and rendered with Mol* via PDBe Mol*, self-hosted here — no third-party CDN, no tracking, no server. The interface is computed from inter-chain heavy-atom contacts, the standard geometric definition of an epitope/paratope footprint. Please cite the RCSB PDB and the structure's primary paper when you use a complex. This tool is part of the free, non-profit Fold Commons project.