Our target is covered by a patent. Can we design a different binder against the same epitope?
In most cases, yes. Antibody patents almost always claim specific sequences, defined by their CDRs or by a percentage identity to them, rather than the epitope itself. When the epitope is not claimed, a different sequence that binds the same site falls outside those claims. We usually get there by starting from the antibody you already know and redesigning around it: the binding mode is already proven, so the work is to move the sequence far enough from the claims while keeping the interaction. Designing from scratch is the second route.
What an antibody patent actually claims
Almost every therapeutic antibody patent claims a molecule, not a target site. The claims list the six CDR sequences, or the full variable domains, often extended by a percentage identity such as "at least 90% identical to SEQ ID NO:1". That is a sequence claim, and it is the kind you can design around.
A smaller number of patents claim any antibody that binds a defined set of contact residues. That is an epitope claim, and it is far broader. If one covers your target, redesigning the sequence will not help, because the claim follows the binding site rather than the molecule. The first question is therefore not scientific but legal: which of the two kinds of claim covers your target?
That question belongs to your patent attorney, and we never substitute for one. What we can say is that in most of the cases brought to us, the epitope itself is not claimed.
Why the epitope is usually free
Broad functional antibody claims have been narrowed on both sides of the Atlantic over the past decade. In the United States, the Supreme Court decision in Amgen v. Sanofi in 2023 raised the bar for claims defined by what an antibody does rather than what it is. At the European Patent Office, epitope claims survive but are held to strict sufficiency requirements. The practical consequence is the same in both jurisdictions: protection has moved back towards sequences.
For anyone holding a blocked programme, that shift is good news: it moves the problem from the target to the molecule. Whether a given design falls outside a given claim is a freedom to operate question, and that belongs to your patent attorney. What we contribute is the technical measurement their opinion rests on: the identity of each candidate to the claimed sequences, computed and documented. We are not describing a loophole. Designing around a patent is one of the purposes the patent system serves, since the invention is published so that others can build something else.
Route one: redesign around the antibody you already have
This is where we start, and in most projects it is where we finish. You give us the published antibody and the structure of the target. We keep the binding mode that already works and redesign the sequence around it, varying the CDRs and the framework until the result sits far enough from the claims.
The advantage is practical. A known antibody tells us which residues carry the interaction and which geometry engages the epitope, so we are not rediscovering how to bind the site. We are only working out how to bind it with a different sequence. That shortens the campaign and raises the hit rate, because the hardest part of the problem is already solved.
It also turns the legal question into a number. If the claim sets a percentage identity threshold, your attorney tells us what margin they want below it, and we design to that figure. Every candidate we deliver comes with its identity to the claimed sequences, computed and documented, so the file is ready if the question is ever raised.
Route two: design from scratch
When the claims are unusually broad, when they cover the epitope rather than a sequence, or when you want a design record with no link at all to the prior art, we generate binders de novo from the target structure alone. Nothing in the pipeline ever sees the patented molecule, which makes the independence of the result straightforward to demonstrate.
This route is more demanding and takes longer, which is why we reach for it second rather than first. It is the right answer when route one cannot put enough distance between your candidate and the claims.
Both routes deliver conventional antibodies or single-domain antibodies, also called nanobodies or VHH, depending on the format your programme needs. A nanobody is often the easier target for this kind of work, because a single domain carries fewer sequence constraints than a paired heavy and light chain.
What we need from you to assess it
A structure of the target, experimental or predicted, and an indication of the epitope you need to engage. If that epitope is defined by a published antibody, the PDB entry is usually enough, and the sequences from the patent itself help us set the distance to aim for.
We then run a free feasibility study that returns a compatibility score for that specific site, telling you whether we can generate credible binders against it before you commit to anything. You never commit to the next step before the previous one has been a success.
Timeline and what you receive
Six weeks from go-ahead to candidate sequences, twelve weeks to binders validated in the laboratory. You receive the sequences, their percentage identity to the claimed prior art, the developability assessment covering aggregation, expression and specificity risks, and the full design record showing how each candidate was produced.
Frequently asked questions
Is designing around a patent legal?
Yes. Courts have long recognised it as one of the ways the patent system benefits the public: the patent publishes the invention precisely so that others can develop alternatives. What matters is that the result falls outside the claims, which is a determination for your patent attorney, not for us.
How different does the sequence have to be?
It depends on the claim. If it specifies a percentage identity, that figure is your target and your attorney will usually ask for a margin below it. If it lists exact CDR sequences, the requirement is different. We design to whatever threshold your counsel sets.
Can you work on a target without the patented antibody?
Yes, through route two. We need the structure of the target and the epitope you want to engage. The campaign takes longer, and the design record carries no link to the prior art.
Who owns the sequences you design?
You do. The sequences, the data and any intellectual property arising from the work belong to the client.
What if the feasibility study comes back negative?
Then we tell you, and the engagement stops there at no cost. A negative answer early is more useful than a failed campaign six months later.
Do you provide a freedom to operate opinion?
No. A freedom to operate opinion is legal work and belongs to a qualified patent attorney. We cover the technical side: the new sequences, their percentage identity to the claimed prior art, and the design record your counsel needs in order to form that opinion. Most clients run the two in parallel, with their attorney setting the threshold and us designing to it.
Working on a target where someone else holds the sequences? Send us the epitope and we will tell you, at no cost, whether we can design around it.
Receive a feasibility study on your antigen - for free