The computational engine for novel radiopharmaceuticals
Radiogenesis is designing the next generation of targeted radioligand therapies
On the Frontier
Radiopharmaceuticals represent one of the most vital frontiers in oncology. Computational drug design can fundamentally transform how these medicines are created.
Proven Efficacy
Early approvals demonstrate that targeted radiotherapeutic drugs, which deliver radiation directly to tumors, can significantly improve patient outcomes in cancers that are otherwise hard to treat.
Human Focused
This modality is only available for 2 tumor types today while many cancer indications remain unsolved. Hundreds of thousands of patients are waiting for better treatment options.
Precision Designed
By combining advances in artificial intelligence, protein engineering, radiochemistry, cancer biology and physics, we accelerate the development of targeted therapies for patients with metastatic cancers.
Approach
Computational Design
Generating and optimizing tailored molecules at scale
We are building an end-to-end computational pipeline for design and optimization of novel radiopharmaceutical drugs.
Our algorithms navigate endless option space to discover diverse and promising candidates beyond nature's blueprint.
Lab in the Loop
Models driven by
experimental wet lab validation
The best scoring molecules are synthesized in the lab, and subjected to extensive validation and testing.
Lab data is fed back into the system to create an iterative learning loop. Canditates are refined towards successful therapeutics.
Technology
Generative design
Create novel molecules by propogating through biological foundation models. Iteratively minimizes a composite set of metrics tailored for targets and epitopes.
Property optimization
Sequence optimization for sidechain packing, energetics, solubility, and stability. Resulting in vectors that express and are highly likely to fold and bind.
Toxicity minimization
Unlock full potential by implementing strategies to reduce nephrotoxicity and hepatotoxicity. This enhances therapeutic efficacy by physicochemical tuning.
Full construct validation
Full construct co-folding and self consistency validation modeling to ensure molecules are well defined and have robust interface geometries.
Isotope agnostic
We are starting with Actinium-225 and Lutetium-177, but our platform is inherently isotope agnostic with the ability to match molecules with various isotopes.
In-silico scoring
Candidates undergo comprehensive scoring, filtering, and ranking by combining deep learning and physics based metrics while ensuring diversity.