Fusiomica | AI-designed proteins for a new era of medicine

Technology - How It Works

Fusiomica is a protein engineering company specialising in the design of de novo mini-protein binders. Leveraging state-of-the-art genAI algorithms, we generate ultra-stable, high-affinity proteins that bind targets inaccessible to antibodies. Unlike natural antibodies (~150 kDa), our binders are 8-15 kDa, enabling rapid tissue penetration, fast clearance and precise targeting. They are produced in E. coli at very high yields and remain folded up to 95 C. A single engineered cysteine permits site-specific conjugation of radionuclides, fluorophores or drugs. The result is a versatile platform that delivers cost-effective imaging probes, biopharmaceuticals, diagnostic and research reagents.

Why de novo binders?

Precision design. Exact binding sites are built in silico for tailored target recognition.
Cost-effective production. Bacterial expression yields monomeric proteins at a fraction of mammalian culture costs.
Enhanced penetration. Small size facilitates access to densely packed tissues and across biological barriers.
Superior stability. AI-optimised hydrophobic cores confer high thermostability and resistance to denaturants.
Simplified IP. Synthetic sequences avoid crowded antibody patent landscapes.
De-immunisation. Designs screened to remove MHC II binding sequences.

How we design

Generative scaffolding. Diffusion models propose shape-complementary folds to target epitopes.
Physics-aware filters. Interface dG, buried surface area, shape complementarity and H-bond satisfaction.
Sequence optimisation. ProteinMPNN-style redesign for expression, stability and solubility.
Manufacturability gates. Cys placement for site-specific conjugation and de-immunisation screens.

Our Offer

We partner with pharmaceutical, biotech and diagnostic organisations to design bespoke binders and deliver experimentally validated constructs. Engagement models include fee-for-service, shared-risk with success fees, and full IP transfer.

Binder Discovery

Computational generation of hundreds to thousands of candidate binders using structure-guided design and deep generative models. Rigorous in-silico screening identifies hits with desired epitope specificity.

Hit to Lead

Expression and purification in E. coli, followed by binding assays (HPLC, SPR, ELISA) to confirm on-target affinity. Lead molecules are affinity-matured to sub-nanomolar range and assessed for stability, solubility and manufacturability.

Functionalisation

Site-specific conjugation of radionuclides, fluorophores or payloads via engineered cysteines or unnatural amino acids. Fusion to Fc domains extends half-life when required. Final constructs are ready for pre-clinical imaging, therapeutic or diagnostic applications.

Case Study: Patent-Pending Probe generated in weeks

To showcase the power of our platform, we pursued a self-funded proof-of-concept programme against a challenging cancer marker ("Target 1"). With a limited bootstrapped budget, our scientists generated multiple binder designs in silico and synthesised the top candidates. Experimental testing confirmed binding for more than 40% of the designs. The selected binder exhibits nanomolar affinity, no unfolding up to 95 C. Production in E. coli yields >100 mg/L of pure protein.

A single cysteine residue, absent from the natural scaffold, allows site-specific attachment of radionuclides, fluorophores or cytotoxins. This binder forms the basis of a European patent application (EP25461534). The entire project was executed 10x cheaper and 80% faster than a comparable antibody discovery campaign.

Project Highlights

Bootstrapped funding. 100% company-funded.
Major unmet need. Target associated with millions of patients per year.
>40% hit rate. High confirmation rate from in-silico design to wet lab.
Patent pending. Composition-of-matter and use claims.
Scalable platform. Lessons generalise across our pipeline.

Experimental Validation

Representative data show purity, binding kinetics and thermostability (target labels redacted).

Purity assessment. Coomassie-stained SDS-PAGE confirms single bands at expected molecular weights.

Monodispersity. Analytical SEC shows a sharp main peak with minimal aggregation.

MicroScale Thermophoresis binding traces

Equilibrium binding. MST traces and binding isotherm demonstrate nanomolar affinity with saturable behaviour.

Applications

GPCR-targeted Binders

GPCR epitopes are buried in flexible extracellular loops and adopt multiple conformations. Our mini-proteins penetrate these loops and can stabilise specific receptor states or block ligand binding.

Diagnostic Reagents

Binders are ideally suited where stability, specificity and ease of production are essential.

CAR-T Cell Engineering

De novo binders offer compact, human-sequence-free recognition domains for cell therapies.

Bispecific and Multispecific Formats

Combining multiple de novo binders enables advanced bispecific, multispecific, and bifunctional molecules.

ADC Formats

Site-specific cysteine engineering enables precise conjugation of cytotoxic payloads for targeted therapies.

Fusion Proteins and Functional Extensions

De novo binders can be fused to additional domains to add new biological functions or pharmacological properties.

Pipeline

We are advancing a diversified portfolio of binder programs, each representing a high-value opportunity in oncology or immunology. Our goal is to develop these de novo binders into next-generation PET probes and theranostics, enabling targeted imaging and therapy.

PET scan panel from Fusiomica presentation summary slide

De novo binders as PET Probes and Theranostics

Our de novo binders are being developed into targeted PET probes that act like a molecular biopsy. These small, stable proteins enable same-day, high-contrast imaging and can be converted into theranostic agents by swapping the imaging isotope for a therapeutic radionuclide.

Whole-body biopsy. Binding probes map antigen expression across all lesions, guiding therapeutic decisions.
Same-day imaging. Small binders clear rapidly from circulation, permitting high-contrast imaging in less than three hours.
Dual-use modality. The same binder can deliver therapy by swapping the imaging isotope for a therapeutic radionuclide.
Reduced invasiveness. Binder-based PET can lower dependence on surgical biopsies and reduce dose burden.

Target 1 Lead Optimisation

Epithelial origin tumours

Target 2 Hit Generation

Splicing isoform, breast

Target 3 Target Selection

Gastric cancer

Note: "Target selection" means prioritisation of new antigens; "Hit generation" is initial binder discovery; "Lead Optimisation" involves affinity, stability and manufacturability engineering.

EU Funded Projects

Contact Us

Ready to explore how our binder technology can benefit your project? Send us a message and we will get back to you shortly.

Email:

Correspondence address: Szlak 77/222, 31-153 Krakow

Deliveries address: MCB UJ, Gronostajowa 7a, 30-387 Krakow | Poland