Relation Therapeutics Unveils MORGAN, an AI Cell Model, with $110M GSK Partnership
Relation Therapeutics has unveiled MORGAN, an AI "foundation model of the cell" designed to predict how human cells respond to drugs and genetic modifications. On the same day, they expanded a collaboration with GSK worth up to $110 million, focused on generating the large-scale cellular data needed to train MORGAN.
The Challenge:
Biological data used to train AI models is scarce, inconsistent, and difficult to obtain.
Relation’s Solution:
They propose manufacturing this data themselves using automated labs that run "superhuman consistent" cellular experiments, generating massive datasets capturing cell responses to various interventions over time.
MORGAN’s Promise:
This model aims to predict disease mechanisms and identify new drug targets directly through computer simulations, potentially revolutionizing drug discovery.
The GSK Collaboration:
GSK will utilize Relation’s data to advance their own drug discovery efforts, focusing on identifying novel therapeutic targets. The deal underscores Big Pharma’s growing interest in investing in AI-driven biological data.
A Crowded but Unproven Field:
Other startups like Chai Discovery and a burgeoning field of AI-bio labs are also developing similar models. While the potential is immense, translating these models into approved drugs remains a significant challenge.