Portfolio

The engine is running on real target pairs.

A design platform is only credible if it has been pointed at something. Four induced-proximity programs are under active design on the TriaX engine: two licensed to TriaX Therapeutics, one a new four-body architecture with its own design pipeline, and one run for an external partner on their targets. They are described here at programme level: the shape of each one, and how far the computation has actually got. Targets, chemistry and candidate structures are held back.

The point of an engine is the second program, and the tenth, not the first.

Four programs from one engine, and the engine is the asset that stays.

Programs on the engine

All four are induced-proximity small molecules that engage more than one non-redundant disease driver with a single designed compound, rather than combining separate drugs. The first two are licensed to TriaX Therapeutics; the third is ours; the fourth belongs to a partner. None is disclosed further at this stage.

PROGRAM 01

Dual-node degrader

A single molecule that drives degradation of two independent nodes a tumour cell depends on at once. The rationale is synthetic lethality: two machineries that a cell cannot compensate for separately, engaged simultaneously, so escape would require rewiring both at the same time.

Area
Oncology
Modality
Induced-proximity degrader
Rights
Licensed · TriaX Therapeutics
PROGRAM 02

Dual-node degrader

The same design logic applied to a second, orthogonal pairing, which is precisely the reuse an engine is supposed to deliver. A second program that costs a fraction of the first is the difference between a platform and a one-off.

Area
Oncology
Modality
Induced-proximity degrader
Rights
Licensed · TriaX Therapeutics
PROGRAM 03

Four-body assembly

A step up in assembly order: three proteins held in one designed complex by a single small molecule, rather than the two-protein pairing every proximity drug on the market today relies on. It is the architecture the engine was built to reach, because a model that reasons about interaction geometry does not care how many bodies are in the picture. Designed on a purpose-built four-body pipeline, described below.

Area
Oncology
Modality
Higher-order induced proximity
Rights
TriaX AI · provisional in preparation
PROGRAM 04

Partner programme

Molecular glue and antigen-directed conjugate pipelines designed on the engine for an external company, against their targets. It is the first programme run as a service rather than licensed into our own pipeline, and it is the cleanest evidence that the engine transfers: a new team, new targets, and the same process end to end.

Area
Partner-directed
Modality
Molecular glue · antigen-directed conjugate
Rights
Partner-owned · provisional in preparation

Stated plainly: the two licensed programs have a complete geometric landscape, every candidate arrangement enumerated, with positive controls passing under the same settings. What is still in progress is the ranking at decision-grade resolution, which is what turns an enumerated landscape into a shortlist worth synthesising. What will be delivered is a ranked set of computationally designed candidates with the provenance behind every judgement, not confirmed binders. Binding data is the gating milestone for both, and we would rather show that boundary than imply a maturity neither program has yet reached. The two newer programs are earlier: design is under way, provisional applications are in preparation, and neither has a completed landscape yet. We will move those markers when the work moves them, not when the plan does.

The programs are the output. The engine is the asset.

TriaX AI designs the molecules and retains the platform and its methods. TriaX Therapeutics takes licensed leads into synthesis and biology, owns composition of matter on the resulting compounds, and runs the programs. Licences are exclusive, field-limited and granted program by program, so the engine stays reusable across future pairings while each program carries its own composition IP. The same engine now also runs on a partner's targets under a service arrangement, where the designs belong to the partner and the methods stay with us. Provisional patent applications covering the platform have been filed; further provisionals covering the four-body architecture and its design pipeline, and the partner-directed designs, are in preparation.

New: a design pipeline for four-body assemblies

Every induced-proximity drug in the clinic brings two proteins together. The engine now designs assemblies with three, held by one molecule, and the pipeline that produces them is new work in its own right.

WHY IT IS HARD

The problem grows faster than the picture

Adding a third protein does not add a third of the difficulty. Every additional body multiplies the arrangements that must be enumerated and adds interfaces that have to be compatible simultaneously, while each new surface is also excluded volume for all the others. A search built around a single pair does not extend to it; it has to be rebuilt.

WHY WE CAN

Geometry does not count bodies

The engine reasons about interaction geometry and flexibility rather than a target family or a fixed complex size, so the step from two bodies to three is a change of scale, not of principle. The same flexed-coordinate treatment, the same excluded-volume accounting, the same gates and controls, applied to a larger assembly.

WHAT IT OPENS

Designs a pair cannot express

A third body can supply a surface neither of the first two has, hold an arrangement that a two-protein complex cannot maintain, or add a selectivity requirement that has to be met at the same time as the others. These are designs with no two-body equivalent, which is the point of building the capability rather than a reason to add complexity.

Where it stands: the pipeline exists and is producing designs; a provisional application covering the architecture and the pipeline is in preparation. There is no binding or cellular data behind it, and we are not claiming any. What is being claimed is a design capability and the process that produces it.

One design axis, three assemblies

Linker differentiation induces structural assembly differences. The same two protein modules and the same pair of anchors, re-bridged at different lengths, produce materially different complexes, so the bridge is not plumbing between two halves, it is the design variable that decides what the assembly becomes.

Surface rendering of a modelled multi-protein induced-proximity assembly
Illustrative. A multi-protein induced-proximity assembly, rendered from our own structural models. Molecular surfaces only, protein identities, chain assignments and the designed compound are not shown. See it assembled →
Cutaway of the assembly with an extended bridge between the two modules
EXTENDED The bridge holds the two modules apart. Proximity by tether.
Cutaway of the assembly with a compacted bridge, module surfaces approaching
COMPACTED A shorter bridge draws the surfaces together. Contact begins.
Cutaway of the assembly at the shortest bridge, drawn at the construction endpoint
FUSED The shortest bridge in the series, at the construction endpoint. Whether two modules can be accommodated this closely is decided per pair.

Cutaway views of the same modelled assembly, rendered from our own structural models under a single fixed camera so the three states are directly comparable. The designed compound is shown in black as a solid volume only, its chemistry is not depicted. Module colours denote the two halves of the assembly and carry no protein assignment.

CONSEQUENCE 01

Tether or interface

An extended bridge holds two modules near each other without bringing them into contact, proximity, but not cooperativity. Shorten it and the surfaces approach, opening the possibility of a genuine interface between the proteins rather than a molecule merely spanning the gap.

CONSEQUENCE 02

Pre-organisation

A long flexible bridge samples an enormous number of shapes, only one of which is productive. A short rigid bridge is locked into a defined geometry before it ever meets the target, a materially smaller entropic price to pay on binding.

CONSEQUENCE 03

Developability

The compact form reaches the same assembly with substantially less molecular bulk. That shows up immediately in the physicochemical profile, which is why bridge length is treated as a developability decision from the first design rather than a property to repair later.

The trade-off, stated: intimate contact demands modules that are sterically compatible with sitting against one another. Where two modules are too large or too awkwardly shaped to do so, compaction is not available and the extended form remains the correct design. In one recent determination the engine searched the complete enumerated landscape for a fusible arrangement and found none, a negative it could state precisely, and act on. The axis is a range to be chosen along deliberately, not a ladder where the tightest option is always the right one.

Designed for breadth

The four active programs sit inside a much wider map of candidate pairings. Because the engine reasons about interaction geometry rather than a single target family, the same process applies across areas, and across both directions of proximity: bringing a protein to a degradation machine, or holding two proteins together to stabilise a complex.

01

Oncology

Two disease drivers engaged at once by one molecule, chosen for non-redundancy so escape has to solve both problems simultaneously.

02

Haematology

Clinically de-risked degradation machinery redirected to co-degrade an additional driver.

03

Neurology

Proximity used to stabilise a protein complex rather than destroy it, the same design problem, run in the opposite direction.

04

Immunology

An established proximity-presenting protein redirected toward new effectors.

Across modalities Degradation Stabilisation Bivalent design Antigen-directed delivery Higher-order assembly
The specifics exist, they sit behind a CDA.

Target pairs, designed structures, ranked shortlists, scoring provenance and programme data are all documented and available to partners under confidentiality. If you are evaluating induced-proximity platforms, tell us the interaction class you care about and we will set up a secure channel before discussing anything sensitive.

Talk to the team →