TCELERATOR™: single-cell functional screening for T-cell engagers
1000s TCE Permutations Ranked by Functional Insights
Designing multispecific T-cell engagers is a combinatorial problem. Each arm, from tumor binders to the CD3 engager, can come in many variants of affinity, epitope and format, and together they produce thousands of possible molecules. Binding assays only answer yes or no, while TCELERATOR™ tells which combination will actually activate T cells against a tumor, by screening and ranking thousands of TCE designs in a single experiment by functional killing, not just binding.
.svg)
.png)
Why us
Our unified engine brings together four proprietary platforms that span the whole development process: PINTRATM for target identification, AFFINISTM for computational binder design and developability tuning, TCELERATORTM for ultra-high-throughput functional screening, and SERISTM for predictive safety assessment. This engine is grounded in real biology validated in our wet lab and built on top of proprietary human tumor datasets spanning more than 10,000 tumor samples and more than 11M single cells. It allows us to significantly cut the development timeline and design multispecific TCEs for patients current therapies leave behind: molecules built to overcome tumor heterogeneity and reach targets current treatments can’t, with safety risks screened out well before the clinic.
Read more in our science explainers: how TCELERATOR screens 10 million single-cell interactions in one run, why function beats binding affinity, and why intracellular targets need TCR-based T-cell engagers.
.png)
A Unified Engine
Proprietary Dataset and Advanced AI Engine to Pinpoint Optimal Targets: PINTRATM
PINTRATM integrates multiple dimensions of data from a large set of tumor and normal samples. It enables us to identify therapeutically relevant, differentiated surface and intracellular targets for solid tumors, providing a therapeutic path for patients with limited or no treatment options.
Computational Protein Design and Optimization to Generate the Best Binders: AFFINISTM
AFFINISTM dissects protein geometry and models molecular interactions computationally to design binders and further optimize them through reinforcement learning, while simultaneously tune other properties to enhance and ensure the molecules' developability.
.svg)
Ultra-high Throughput Functional Screening to Find the Therapeutic Leads: TCELERATORTM
TCELERATORTM functionally profiles more than 10 million single-cell interactions at once. Where conventional discovery tests binders sequentially over months, it identifies functional leads in a single experiment. The platform now screens TCE binders and formats directly, ranking binders and molecular geometries by their functions rather than by binding affinity alone.
AI-Driven Deep Molecular Scanning to Predict Potential Adverse Reactivity for Clinical Safety: SERISTM
SERISTM computationally screens epitope interactions of the TCE candidates with normal human tissues to look for potential cross-reactivity. This approach has enabled us to successfully predict previously reported clinical toxicities to the tissue/organ, providing a powerful safeguard for our TCE design process.
.svg)
