{"slug": "gsk-and-relation-expand-ai-drug-discovery-collaboration", "title": "GSK and Relation Expand AI Drug Discovery Collaboration", "summary": "GSK and Relation Therapeutics entered a research collaboration on July 30 worth up to $110 million to generate human cellular perturbation datasets and train AI foundation models for drug-target discovery. The collaboration extends an earlier relationship focused on fibrotic diseases and osteoarthritis, and includes Relation's MORGAN platform. Relation CEO David Roblin said the work aims to deepen understanding of disease biology through richer data for therapeutic target discovery.", "body_md": "# GSK and Relation Expand AI Drug Discovery Collaboration\n\nGSK and Relation Therapeutics entered a research collaboration on July 30 worth up to $110 million to generate human cellular datasets and train AI models for therapeutic target discovery. Reuters reports that the work will use data on cellular responses to genetic and drug interventions, including for Relation's MORGAN foundation-model platform.\n\nGSK and Relation Therapeutics entered a research collaboration worth up to **$110 million** to generate human cellular perturbation datasets and train AI foundation models for drug-target discovery. Reuters reported on July 30 that the collaboration extends an earlier relationship between the companies focused on fibrotic diseases and osteoarthritis.\n\nUnder the agreement, Relation will generate large-scale datasets measuring how human cells respond to genetic and pharmacological interventions. Reuters reported that the data will be used to train foundation models, including Relation's MORGAN platform. Fierce Biotech reported that Relation is eligible for the $110 million through upfront and success-based milestone payments.\n\n### Data generation and model training\n\nPerturbation data captures biological measurements before and after an intervention, such as a gene edit or drug treatment. According to Fierce Biotech, the companies intend to use the resulting data to train models aimed at identifying new therapeutic opportunities.\n\nDDW reported that the work will use integrated automation to produce time-resolved perturbation data with multi-omics readouts at scale. Multi-omics datasets can combine measurements across biological layers, such as gene regulation and other molecular activity, rather than representing cells through a single assay.\n\nRelation CEO David Roblin said in a statement reported by DDW: \"Deepening our understanding of the underlying biology of disease starts with richer data, to be used in models that can give us greater confidence in the discovery of therapeutic targets that can ultimately yield medicines.\"\n\nBioXconomy reported that Relation unveiled MORGAN, short for Multi-Omic Regulatory Genomics using Artificial Neural Networks, on the same day as the GSK agreement. Roblin described MORGAN to BioXconomy as a model for cellular perturbation responses across disease contexts, trained using petascale, high-resolution multi-omic datasets produced through automated laboratories.\n\n### Why the dataset is central\n\nThe collaboration places the data-collection layer alongside model development rather than treating foundation-model training as a standalone software problem. According to Relation's description reported by Fierce Biotech, MORGAN is designed to predict how cells respond to perturbations across disease contexts.\n\nFor ML practitioners working in biology, this reflects a recurring constraint in scientific foundation models: model capability depends heavily on the coverage, reproducibility, and experimental design of the training corpus. Time-resolved intervention data can be particularly useful because it links molecular measurements to an explicit intervention, although the practical value of a model remains dependent on validation in relevant biological systems.\n\nThe companies have not disclosed specific disease targets for the new collaboration, Fierce Biotech reported. GSK has been increasing research-and-development investment and expanding AI use in drug discovery, Reuters reported, while Relation's work spans areas including immunology, metabolic disease, and bone disease, according to Fierce Biotech.\n\nThe agreement provides a concrete example of AI drug-discovery partnerships centered on proprietary experimental data. Comparable efforts in the sector often combine automated wet-lab generation, multi-omic measurement, and machine learning because public biological datasets may not contain the intervention depth or consistency needed to train models for target prioritization.\n\n## Key Points\n\n- 1GSK and Relation's deal supports human cellular perturbation datasets, connecting automated experimental biology directly to foundation-model training for target discovery.\n- 2The collaboration uses multi-omics, time-resolved intervention data, which can provide explicit intervention structure beyond static observational biological datasets.\n- 3Across AI drug discovery, proprietary dataset quality and biological validation often determine whether foundation-model predictions become usable target hypotheses.\n\n## Scoring Rationale\n\nThe $110 million collaboration is a notable deployment of foundation-model methods in pharmaceutical target discovery, with an emphasis on generating specialized biological training data. It is relevant to ML practitioners building scientific models, though the sources report no model performance or validated discovery outcomes.\n\n## Sources\n\nPublic references used for this report.\n\nPractice interview problems based on real data\n\n1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.\n\n[Try 250 free problems](/problems)", "url": "https://wpnews.pro/news/gsk-and-relation-expand-ai-drug-discovery-collaboration", "canonical_source": "https://letsdatascience.com/news/gsk-and-relation-expand-ai-drug-discovery-collaboration-087ca2a1", "published_at": "2026-08-03 10:00:00+00:00", "updated_at": "2026-08-03 11:00:02.993910+00:00", "lang": "en", "topics": ["artificial-intelligence", "machine-learning", "ai-research", "ai-products"], "entities": ["GSK", "Relation Therapeutics", "MORGAN", "David Roblin", "Reuters", "Fierce Biotech", "DDW", "BioXconomy"], "alternates": {"html": "https://wpnews.pro/news/gsk-and-relation-expand-ai-drug-discovery-collaboration", "markdown": "https://wpnews.pro/news/gsk-and-relation-expand-ai-drug-discovery-collaboration.md", "text": "https://wpnews.pro/news/gsk-and-relation-expand-ai-drug-discovery-collaboration.txt", "jsonld": "https://wpnews.pro/news/gsk-and-relation-expand-ai-drug-discovery-collaboration.jsonld"}}