Key points from article :
GSK is strengthening its investment in artificial intelligence-driven drug discovery through an expanded partnership with London-based biotech company Relation. The agreement, worth up to $110 million, will focus on generating large-scale biological datasets that could help researchers identify promising drug targets with greater accuracy.
The collaboration reflects a growing trend across the pharmaceutical industry, where companies are combining advanced AI models with detailed human biological data to improve the early stages of drug development.
Under the agreement, Relation will create large datasets showing how human cells react when they are exposed to genetic changes or drug interventions. These measurements, known as cellular perturbation data, allow scientists to observe how biological systems respond when specific genes or cellular pathways are altered.
The resulting data will then be used to train AI foundation models designed to better understand disease biology and uncover potential therapeutic opportunities.
One of the models involved will be Relation's MORGAN platform, short for Multi-Omic Regulatory Genomics using Artificial Neural Networks. MORGAN is designed to predict how cells will respond to different biological interventions across a range of disease settings.
For drug developers, this type of technology could address one of the industry's biggest challenges: selecting the right biological targets.
Developing a new medicine is an expensive and lengthy process, and many experimental drugs ultimately fail during clinical development. By combining high-quality human biological data with predictive AI models, GSK and Relation hope to build greater confidence around potential targets before significant resources are committed to developing new medicines.
Relation CEO David Roblin emphasized the importance of data in understanding complex diseases. According to the company, generating information from physiologically relevant human disease systems could reveal biological mechanisms that traditional approaches might overlook.
Financially, Relation could receive as much as $110 million through a combination of upfront and success-based milestone payments. The companies have not disclosed which specific diseases will be targeted under the expanded collaboration. Relation currently conducts research in areas including immunology, metabolic disorders and bone diseases.
The agreement also deepens an existing relationship between the two companies. GSK and Relation previously signed two collaborations in 2024 focused on discovering and validating new drug targets for fibrotic diseases and osteoarthritis.
Relation has also attracted interest from other major pharmaceutical companies. In December, it entered a separate collaboration with Novartis focused on discovering targets for atopic diseases. That agreement included $55 million in upfront payments, research funding and an equity investment, with Relation potentially eligible for additional milestone payments.
The latest GSK partnership highlights how AI drug discovery is moving beyond simply developing better algorithms. Pharmaceutical companies increasingly need large amounts of relevant, high-quality biological data to make those algorithms useful.
If successful, the GSK-Relation collaboration could demonstrate how combining human cellular experiments with AI models can make early drug discovery more precise. Rather than replacing laboratory science, AI could become a powerful tool for interpreting increasingly complex biological data—and potentially help researchers choose better drug targets before they enter the costly stages of clinical development.
