Data Scientist @ Datwave
Director Translational Data Science @ Alentis Therapeutics
In the world of medicine, there is a traditional “waiting period”. Researchers can observe how a drug behaves in preclinical models, but they cannot see its effect on a human organ until clinical trials are already underway.
To bridge this gap, Datwave, in partnership with Alentis Therapeutics and Google Cloud, created the FibroTransform Pipeline. This AI‑powered tool learns the visual signatures of response from preclinical studies and projects those tissue‑level changes onto untreated human biopsy images. For the first time, scientists can preview how a biopsy might look post‑treatment and before dosing begins.
Developing new treatments for serious conditions like cancer and organ fibrosis is a race against time. Alentis Therapeutics is developing lixudebart, a first‑in‑class antibody targeting exposed Claudin‑1 (CLDN1) in fibrotic tissues.
Translating successful preclinical data to human patients is always a hurdle. Because human and preclinical images differ—across species, acquisition, magnification, and color (staining)—envisioning a potential inter‑species outcome felt like a missing piece of the puzzle.
Datwave built a translational “bridge” using advanced Generative AI. Using CycleGAN for unpaired image‑to‑image translation, our model learns response‑linked morphology in preclinical animal tissue from hundreds of images. This smart architecture acts as a “biological translator,” rendering a synthetic counterpart on untreated human biopsy images. Running on Google Cloud, it processes high‑resolution medical images to simulate plausible tissue‑level outcomes.
The results were a promising breakthrough for medical simulation: the AI‑generated images were visually realistic and medically relevant. They captured key anti‑fibrotic patterns consistent with preclinical findings, providing a tissue‑level visualization of what a human biopsy might look like after treatment. This project showed that AI can turn cross‑species histology into a practical decision‑support asset, accelerating hypothesis generation and study design.
Director Translational Data Science @ Alentis Therapeutics
Collaborating with Datwave was smooth and effective. We aligned quickly on goals, and their expertise drove the development of meaningful tools shaping the future of translational models in clinical development.
Please note: the contact form is displayed only after you accept cookie policy.
