Generating human-cell data to train AI models for drug discovery
Drug-discovery models need data showing how human cells respond to changes, not just snapshots of their state. Common datasets use young or cancer-derived cells that may poorly represent age-related disease. These mismatches make it harder to learn disease mechanisms and predict responses to interventions.
CellCodex generates data by changing genes and applying compounds to disease-relevant human cells. It measures individual cells through molecular profiling and imaging, supplying data to train AI models that prioritise drug targets and experiments.
What stands out about CellCodex
Conducted genome-wide screens in human dopamine neurons that pointed to a novel pathway in its Parkinson's disease programme.
Co-founded by Grant Belgard, who previously founded The Bioinformatics CRO and brings single-cell genomics expertise.
Co-founded by Manos Metzakopian, PhD, a former bit.bio CSO and UK Dementia Research Institute group leader in Cambridge.
Co-founder Tom Weaver founded and led PetMedix, which was acquired by Zoetis in 2023.
Co-founders previously held leadership roles at bit.bio (Manos Metzakopian, CSO) and Scorpion Therapeutics (Ravi Moorthy, VP Communications).
Co-founder Grant Belgard earned his DPhil at Oxford and subsequently worked there as a research scientist.
Get a curated pipeline of healthcare and life sciences startups aligned with your investment thesis.