Why biological data matters more in AI drug discovery
$110M. GSK and Relation Therapeutics just bet that biological data is the constraint in AI drug discovery, not model capability.
Why it matters
A major pharma company is paying for curated biological datasets to train AI models — signaling that frontier capability means nothing without domain-specific training data. This shapes where AI drug-discovery budgets flow.
The key facts
9 to knowGSK-Relation Therapeutics collaboration worth up to $110 million
Focus on large-scale datasets measuring human cell response to genetic changes and drug interventions
Data generated to train AI models for drug discovery
Expands existing work between the two companies
Reflects shift from model capability to training data as the bottleneck in biotech AI
Focus: generating large-scale datasets measuring human cell responses to genetic changes and drug interventions
Data intended to train AI models for drug discovery
Expansion of existing partnership between the companies
Pharmaceutical industry bet on data as the AI bottleneck, not model capability
Go to the source
AI Newsartificialintelligence-news.com
Publisher excerpt: GSK has entered into a research collaboration with British biotechnology company Relation Therapeutics worth up to $110 million, expanding the companies’ existing work in AI-assisted drug discovery. Under the agreement, Relation will generate large-scale datasets measuring how human cells respond…