LabGenius Therapeutics provides the EVA™ platform, a closed‑loop discovery engine that integrates machine‑learning‑based in silico design with automated high‑throughput cell‑based screening to create multispecific therapeutic antibodies. The system employs multi‑objective Bayesian optimization to simultaneously maximize potency, selectivity, and developability across formats such as bispecifics, trispecifics, ADCs, and cell engagers, delivering IND‑ready candidates in under six weeks.
Funding
Funding not disclosed
Founders
Product
Problem
Designing multispecific therapeutic antibodies is hampered by a non‑intuitive relationship between sequence, structure, and functional activity, making conventional engineering inefficient and leading to molecules with suboptimal efficacy or on‑target, off‑tumor toxicity in solid‑tumor indications.
Solution
LabGenius Therapeutics addresses this gap with the EVA™ platform, a closed‑loop discovery engine that fuses machine‑learning‑driven in silico design with automated high‑throughput cell‑based screening. Multi‑objective Bayesian optimization evaluates millions of design candidates across functional potency, selectivity, and developability metrics, enabling rapid co‑optimization of complex formats such as bispecifics, trispecifics, ADCs, and cell‑engagers. The platform delivers selectivity‑enhanced antibodies that achieve complete on/off killing of tumor cells while sparing healthy tissue, compressing the design‑build‑test‑learn cycle to under six weeks. This capability accelerates IND‑enabling candidate generation for solid‑tumor programs without the need for extensive manual iteration.
Target Audience
Primary customers are pharmaceutical and biotech R&D organizations developing antibody‑based therapeutics for solid‑tumor indications, as well as contract research organizations seeking accelerated lead optimization of multispecific formats.
Features
- Proprietary ML models trained on ML‑grade experimental data to predict affinity, epitope coverage, and developability attributes.
- Multi‑objective Bayesian optimization that simultaneously maximizes potency, avidity‑driven selectivity, and manufacturability.
- Automated modular cloning, sequence verification, mammalian expression, and purification pipeline capable of processing up to 2,300 multispecific constructs per cycle.
- High‑throughput disease‑relevant cell‑based functional assays integrated with developability screens (stability, aggregation, expression yield).
- Modality‑agnostic workflow supporting bispecific T‑cell engagers, trispecific formats, ADCs, NK‑cell engagers, and radioconjugates.
- End‑to‑end data pipeline with cloud‑hosted analytics, version‑controlled design libraries, and secure data storage compliant with industry standards.
- Avidity‑based selectivity engine that exploits differential antigen density to achieve >10⁴‑fold therapeutic windows in vitro.