November 19, 2026
Antibody patent claims: review the evidence for breadth
Review antibody sequences, binding and function assays, and representative examples to connect proposed claim breadth to the disclosure supporting it.

How broadly can you claim an antibody invention? Before choosing antibody patent claim breadth, inspect 3 connected evidence sets: the sequences defining the antibodies, the assays supporting each claimed property, and examples representing the variation you want to claim. Connect that evidence to the application’s disclosure so you can identify supported claim options and unresolved gaps.
This review addresses U.S. disclosure support and targeted European Patent Office (EPO) antibody-claim requirements.
Define each claim-scope proposal separately
Start with the commercial objective: protection for the lead antibody, a family of variants, or antibodies sharing a binding or biological property. Then write the actual scope alternatives. “Broad antibody protection” leaves too many scientific choices hidden.
A claim might specify paired heavy- and light-chain variable-region sequences, particular complementarity-determining regions (CDRs), sequence variants retaining a defined activity, or antibodies with specified functional properties. CDRs are regions within the antibody variable domains involved in antigen binding. Each change in the proposed definition creates a corresponding support question.
For U.S. claims, keep two inquiries distinct. Written description asks whether the disclosure conveys possession of the claimed invention. Enablement asks whether it teaches a skilled person to make and use the claimed scope without undue experimentation. The United States Patent and Trademark Office (USPTO) treats these as separate disclosure requirements ↗. The broader Section 112 review ↗ supplies the general framework; the antibody-specific work is connecting each scope proposal to its technical evidence.
Keep earlier publications and sequence-search results in the prior-art record. Use the patent sequence-search workflow ↗ for retrieval and alignment evidence. For claim support, identify what your own application teaches, with a paragraph, sequence identifier, example, or figure for each proposition.
Inspect sequence identity at the chain and residue level
Build an antibody-by-antibody sequence record before treating a collection of sequences as a family. For each disclosed antibody, capture:
- Molecular identity: antibody name, format, heavy/light pairing, variable-region boundaries, and relevant constant-region or engineered features.
- Sequence locations: the amino acid sequences, their sequence identifiers such as SEQ ID NO: 1, and the application passages connecting those identifiers to the antibody actually made or described.
- CDR assignments: each CDR sequence, the numbering convention used, and its location within the corresponding variable region.
- Permitted variation: specified substitutions, insertions, deletions, framework changes, and any proposed identity range.
- Evidence status: which construct was produced, which was assayed, and which remains a proposed embodiment.
Treat chain pairing as part of the evidence. If the record describes heavy chain H1 paired with light chain L1, locate the teaching for any proposed H1/L2 combination. Likewise, connect an assay result to the tested format before extending the same proposition to a fragment or a differently engineered antibody.
EPO practice makes the CDR mapping particularly concrete. Under its structural-definition guidance ↗, an antibody’s structural definition must include the CDR sequences required for antigen binding. For immunoglobulin G (IgG), that ordinarily means all 6 CDRs. Omitting a CDR requires experimental evidence that it does not interact with the antigen. When identifying CDRs through larger chain sequences, include the numbering scheme. These are Article 84 claim-definition requirements; assess sufficient disclosure separately.
For a percentage-identity proposal, specify the reference sequences, comparison method, relevant regions, and allowed changes. Then ask which disclosed evidence supports retained activity across those changes. The EPO permits combined structural and functional definitions ↗ using less than 100% sequence identity. Its guidance supplies no percentage that automatically supports a variant family.
Match every functional limitation to its assay
Separate target binding, binding affinity, competition with another antibody, and biological effects such as receptor blocking. Assign a distinct evidence entry to every property in the proposed claim, including properties joined by “and.”
For each assay, inspect the protocol and underlying results alongside the summary slide. A useful entry records:
- Tested materials: the exact antibody construct and pairing, antigen sequence or construct, species, and presentation, such as soluble protein or a cell-surface target.
- Method and conditions: assay type, reagent concentrations, temperature, incubation time, instrument or analysis method where relevant, and the protocol version.
- Readout and boundary: what was measured, units, the claimed threshold, and how a result is assigned inside or outside that threshold.
- Controls and comparators: negative and positive controls, reference-antibody identity, and the conditions used for any comparison.
- Results: individual measurements or curves, repeat experiments, variability, and any failed or inconclusive runs relevant to the claimed property.
For example, a claim combining target binding with receptor blocking calls for teaching supporting both properties. An entry reporting binding should direct counsel to the separate blocking evidence, or identify blocking as an open question. An epitope is the antigen region recognized by an antibody. If a competition assay supports proposed epitope language, record exactly what the experiment measured and the basis for any residue-level characterization.
The EPO’s functional-claim guidance ↗ requires an application to enable further antibodies having the claimed property without undue burden. It also calls for a functional definition whose boundaries can be determined easily and unambiguously, normally with the relevant measurement-method characteristics in the claim.
Your practical question is therefore twofold: can the evidence identify antibodies meeting the proposed boundary, and does the disclosure teach how to obtain further members? Put the assay definition and the generation or selection teaching in separate fields so a positive result cannot obscure an unresolved production question.
Examine what the examples represent
Organize examples by the variation relevant to the proposed claims: different CDR sequences, frameworks, pairings, formats, or functional profiles. Repeated measurements strengthen the record for a tested construct. To assess broader scope, identify which additional structural or functional differences the examples address.
For a U.S. genus claim, meaning a claim to a class of antibodies, USPTO guidance permits reliance on representative species, identifying characteristics, or a combination. Functional characteristics require a known or disclosed structure-function correlation. Its representative-species guidance ↗ focuses on variation within the claimed class. For your evidence record, assess that variation rather than applying a fixed example quota.
Ask the inventors to explain the proposed generalization in scientific terms:
- Which shared features identify the claimed family?
- What evidence connects those features to each claimed activity?
- Which substitutions or combinations were actually evaluated?
- What failures or activity losses reveal a boundary?
- How would a skilled person obtain additional members, and what screening or optimization remains?
The Supreme Court’s 2023 decision in Amgen v. Sanofi ↗ illustrates the distinction. The patents disclosed 26 sequence-identified antibodies, yet the Court held the asserted broad functional claims insufficiently enabled. The proposed routes to further members left substantial trial-and-error discovery. The Court allowed reasonable experimentation and recognized that a shared quality can sometimes support teaching an entire class. For your review, examine the teaching that extends beyond the examples as closely as the examples themselves.
The EPO’s sufficiency guidance ↗ similarly connects broad scope to enabling teaching across that scope. One or a few examples can suffice where the disclosure and common general knowledge enable substantially any claimed embodiment without undue burden. Assess the evidence for the particular claim and jurisdiction.
Apply the review to a synthetic antibody family
Consider this fictional October 2026 filing packet for antibodies against Target Q:
- Sequence schedule S1: paired variable-region sequences are assigned to QX-1 (SEQ ID NOs: 1 and 2), QX-2 (3 and 4), and QX-3 (5 and 6). QX-1 and QX-2 have the same 6 CDRs in 2 frameworks, with an identified numbering scheme. QX-3 changes one heavy-chain CDR3 residue. Two further variants are proposed.
- Example 1: expression and purification methods identify the constructs produced. All 3 use the same IgG format.
- Example 2: surface plasmon resonance measurements show that all 3 produced antibodies bind soluble human Target Q at 25°C. The protocol and individual binding curves are included.
- Example 3: a cell-surface ligand-binding assay shows that QX-1 and QX-2 block Target Q’s interaction with its ligand. QX-3 has no blocking result. The 2 proposed variants have no experimental results.
Counsel can turn that packet into a claim-support review record. Add exact source locations as the specification develops, preserving the distinction between a demonstrated result and a proposed embodiment.
| Proposed claim scope | Evidence in the synthetic packet | Unresolved support question | Counsel’s next action |
|---|---|---|---|
| QX-1’s specified heavy/light variable-region pair with binding and blocking properties | S1 identifies the pair; Examples 1-3 describe production and both activities | Does the application consistently connect the sequences, tested construct, methods, and results? | Reconcile identifiers and carry the linked technical teaching into the specification. Assess the complete claim. |
| Antibodies with QX-1’s 6 CDRs across other frameworks or formats | QX-1 and QX-2 provide 2 frameworks in the disclosed format | What supports the additional frameworks or formats encompassed by the wording? | Identify the shared-feature rationale and relevant alternatives. Compare a more specifically defined family with the broader proposal. |
| Antibodies having at least 95% identity to the reference variable domains and retaining binding and blocking | 3 binding results, 2 blocking results, and proposed variants | Which allowed changes preserve both properties, and how does the disclosure teach further members? | Distinguish tested substitutions from predictions. Identify missing functional evidence and generation guidance before selecting the variant language. |
QX-3’s binding result leaves a specific blocking question open; adding more binding measurements would leave that question unresolved.
Use the record to choose the next substantive step: refine the wording, locate existing technical evidence, obtain additional inventor explanation, or plan experiments addressing a defined gap. Counsel can then compare the supported options against the protection the program needs.
Close the record against the application you will file
Before settling claim breadth, confirm that the application contains the sequence relationships, assay methods, examples, and explanatory teaching relied on in the review. An internal laboratory record needs a corresponding application location where its teaching is necessary to support the claim.
Keep actual results distinct from proposed or prophetic examples. U.S. enablement practice permits applications without working examples in appropriate circumstances and evaluates enablement at filing. Later evidence cannot supply missing filing-date enablement. The USPTO’s enablement guidance ↗ addresses these distinctions.
For each complete claim option, record counsel’s disposition and the remaining evidence question. That gives the drafting team a concrete instruction and gives the scientific team a specific question to resolve.
Frequently asked questions
How many antibody examples support a genus claim?
There is no fixed number. Under the USPTO’s written-description guidance ↗, species must represent relevant variation, or identifying characteristics must otherwise convey possession of the genus. Enablement separately turns on teaching the claimed scope without undue experimentation. More examples help when they address differences that matter to the proposed breadth.
Do EPO antibody claims need all six CDRs?
For an IgG defined structurally, ordinarily yes. The EPO requires the CDR sequences needed for antigen binding, normally all 6. Fewer can be claimed when experiments show that an omitted CDR does not interact with the antigen. A larger chain-sequence reference also needs a numbering scheme under the EPO’s structural-definition guidance ↗.
Can a percentage-identity claim be supported without testing every variant?
Yes, when the disclosure supports the claimed breadth and teaches how to make and use it without undue experimentation. Amgen v. Sanofi ↗ permits reasonable experimentation in U.S. practice. The EPO also permits combined structural-functional definitions ↗ below 100% identity. Define the reference sequences, permitted changes, retained function, and teaching for obtaining further members.
Where Patlytics fits
Patlytics supports inventor disclosure, application drafting, specification-support review, and the separate prior-art review:
- Invention Disclosures (IDF) turns inventor materials, including slides and figures, into a reviewable disclosure.
- Patent Drafting ↗ supports proposed claims and specification text. Drafting runs in a native Word editor with tracked changes and comments, and claim numbering holds.
- §112 Check flags written description, enablement, and definiteness issues, with citations back to the specification.
- Patent & NPL Search ↗ supports the separate prior-art review across patents and non-patent literature (NPL).
Life-sciences drafting inputs include sequence uploads, ChemDraw input, a structure editor, image-to-SMILES conversion, and Markush/R-group generation for generic chemical structures. SMILES means Simplified Molecular Input Line Entry System, a notation for molecular structures.
Attorneys retain responsibility for reviewing the scientific evidence, application support, and final claim-scope decisions.
Bring the antibody disclosure and proposed claim alternatives into a reviewed drafting process with Patlytics. Book a demo →
See it on your own patents
Reduce cycle times. Increase margins. Deliver winning IP outcomes.


















































