Relation has launched MORGAN (Multi-Omic Regulatory Genomics using Artificial Neural Networks), a flagship cellular foundation model, designed to transform how new biology is understood and medicines are discovered.
MORGAN is a general-purpose model which can be applied across cell types and applied to specific disease areas.
Each tissue-specific MORGAN model will focus on a therapeutically important cell type that plays a central role in disease biology.
Built using large-scale, high-resolution multi-omic perturbation data, these models are designed to predict how human cells respond to genetic and pharmacological interventions, providing unprecedented insight into disease mechanisms and helping identify and validate new therapeutic opportunities.
MORGAN combines both frontier-scale computation and an entirely new approach to experimental data generation. Relation will generate petascale multi-omic perturbation datasets with the consistency, quality and scale required to train MORGAN.
The data will arise from automated laboratories designed specifically for high-throughput cellular perturbation experiments with superhuman consistency, enabling data generation beyond what is achievable through conventional laboratory approaches.
Together, MORGAN and the experimental platform that powers it, represent the most significant effort to build therapeutically relevant models of the cell. By combining industrial-scale data generation with state-of-the-art machine learning, Relation is establishing the scientific infrastructure needed to accelerate the discovery of future medicines.
