An antibody patent search looks for patents and publications that disclose or claim an antibody. A claim may define the antibody by its sequence, its complementarity-determining regions (CDRs), its epitope or its function, and each definition needs its own search method. The method builds on a standard biological sequence search.
How antibody claims are written
The Guidelines for Examination of the European Patent Office (EPO) list the main ways to define an antibody in a claim (G-II, 6.1, checked November 2026):
- its amino acid sequences, or the nucleic acid sequences encoding it;
- the target antigen, alone or with further functional features;
- functional and structural features together;
- the production process, or the hybridoma producing it.
Each format points to different prior art, so an antibody patent search is planned claim type by claim type.
| Claim type | Main search method |
|---|---|
| CDR or variable domain sequence | Sequence alignment, with each CDR as a separate query |
| Target antigen | Text and classification search on the antigen and its synonyms |
| Epitope | Antigen sequence search, then reading of mapping data |
| Function | Text search on the antigen with functional terms and assay names |
Searching CDR sequences in an antibody patent search
In an immunoglobulin G (IgG), the binding region has one heavy and one light chain variable domain, each with 3 CDRs. Under EPO practice, a claim defining an IgG by structure must normally recite all 6 CDRs (G-II, 6.1.1).
CDRs are short, so a search on the full variable domain can rank an exact CDR match below longer, weaker alignments. The searcher therefore runs each CDR as its own query and also searches the 6 CDRs as a set. A biological sequence search records the settings used for these short queries.
CDR boundaries also depend on the numbering scheme. The EPO Guidelines name Kabat, Chothia and IMGT as examples, and the schemes can place CDR boundaries at different residues. The searcher extracts the CDRs under each scheme that the relevant documents use.
Variable domains and sequence variants
Many claims recite both variable domains at a stated percent identity, combined with a function. Section G-II, 6.1.4 of the EPO Guidelines accepts sequences with less than 100% identity when they are combined with a clear functional feature. In an antibody patent search, the threshold is set below the lowest identity claimed, and hits near it are read by hand, as the biological sequence search guide describes.
Antibody sequences also appear in sources outside patent sequence listings. The Patent and Literature Antibody Database (PLAbDab) holds over 150,000 paired antibody sequences taken from patents and academic papers.
Searching epitope claims
An epitope claim defines the antibody by the part of the antigen it binds, so the search moves to the antigen. The searcher aligns the antigen sequence and finds documents on antibodies to it. Their epitope mapping data usually sits in tables, figures and examples rather than in the listing.
Functional antibody claims: the gap a sequence search leaves
A claim to any antibody that binds antigen X and blocks its receptor has no sequence to align. Under section G-II, 6.1.3, unless there is an indication to the contrary, the EPO assumes that a prior-art antibody binding the same antigen has the claimed functional properties. For patentability or invalidity work, every earlier antibody to that antigen is a candidate reference. Claims defined by competition with a reference antibody first disclosed in the application normally cannot be completely searched, according to the EPO.
For freedom to operate (FTO), the risk runs the other way. A functional claim can cover a product whose sequence appears in no patent. An FTO search for an antibody therefore adds antigen and function searching to the sequence work described in the guide to pharma FTO searches for biologics.
In Amgen Inc. v. Sanofi, decided in 2023, the US Supreme Court held unanimously that claims to a whole genus of antibodies defined by their function were not enabled. The specification had given the sequences of only 26 antibodies. However, functional claims granted earlier may still be in force.
What an antibody patent search report contains
An antibody patent search report usually contains:
- the queries, including CDRs under each numbering scheme;
- databases, parameters and search dates;
- results grouped by claim type;
- where each match appears: claims, description or listing only;
- notes on functional claims that alignment cannot settle;
- legal status, where the search is for FTO.
Scoping an antibody patent search
The scope of an antibody patent search depends on the purpose, the claim types expected and the countries of interest. Before it starts, the search firm usually needs:
- the heavy and light chain sequences;
- the CDR numbering scheme used;
- the target antigen.
For an antibody patent search, see our biological sequence search service or contact us.
This guide is general information and is not legal advice.
Frequently Asked Questions
What is a CDR search in patents?
A CDR search aligns each complementarity-determining region of an antibody against patent and literature sequences as a separate short query. It finds documents that disclose or claim the same binding regions, which a search on the whole variable domain can rank too low.
Which CDR numbering scheme is used in patent searches?
No single scheme is used, because patents define CDRs under Kabat, Chothia, IMGT or other schemes. The boundaries differ, so one residue can fall inside a CDR under one scheme and outside it under another. The search covers each scheme the relevant documents use.
Can a new antibody against a known antigen be patented?
It can at the EPO when the application shows a surprising technical effect, such as improved stability, or when there was no reasonable expectation of success. Overcoming difficulties in making the antibody can also count, but a structural difference alone, such as a new sequence, does not (G-II, 6.2).
Where are antibody sequences published outside patents?
Antibody sequences appear in research articles, in public databases such as GenBank and in curated collections such as PLAbDab. Searchers add these sources because a journal article or a database entry can be prior art against a later patent claim.
Why can a sequence search miss a relevant antibody patent?
A sequence search misses claims that define the antibody only by its antigen, epitope or function, because these claims recite no antibody sequence. It can also miss older documents that show sequences only in figures. Text searching on the antigen and its functions closes most of this gap.