SEATTLE and VIENNA — Parse Biosciences and Graph Therapeutics said they have entered into a strategic partnership to create one of the largest and most comprehensive functional immune cell perturbation atlases, combining artificial intelligence with single-cell biology to improve drug discovery and development for immune-mediated diseases.
The collaboration will integrate Graph’s lab-in-the-loop platform with Parse’s GigaLab to profile hundreds of millions of cells from patients with immune diseases under systematic perturbation. The companies said the effort is designed to make the immune system’s highly dynamic behavior accessible to AI-first drug discovery, with the goal of reducing risk and accelerating the development of new therapies.
Autoimmune and immune-mediated diseases are driven by context- and patient-specific immune cell responses, which often makes it difficult to identify drug targets and predict clinical outcomes. Parse and Graph said their combined approach addresses this challenge by pairing Graph’s patient-derived disease models with Parse’s scalable whole-transcriptome single-cell technology, enabling large-scale analysis of human immune cells and the interactions that drive disease. The companies said this strategy could help prevent costly late-stage clinical failures and reduce overall drug development costs.
Graph’s platform uses primary patient cell assays alongside an iterative active-learning framework to systematically select and test perturbations across a wide range of disease-relevant conditions. Rather than relying on simplified or static models, the approach is designed to distinguish promising therapeutic hypotheses from nonviable ones earlier in development. Each experimental cycle feeds into the next, creating a compounding knowledge base intended to accelerate future discoveries.
Once Graph’s platform identifies which disease contexts and perturbations to study, Parse’s GigaLab will generate large-scale single-cell datasets using its Evercode technology, delivering high-throughput data with speed and consistency.
“After a decade of evolving impactful AI drug discovery platforms, we’ve learned that closing the gap between prediction and clinical reality demands active investment in biological context,” said Gregory Vladimer, PhD, chief executive officer and co-founder of Graph Therapeutics. “This partnership treats data as strategic infrastructure: every experiment is designed to reduce biological uncertainty, every validation compounds institutional knowledge, and every discovery accelerates the next. When you invest in fit-for-purpose, clinically relevant data generation at the discovery stage, you fundamentally change the economics and success rates of drug development.”
“Graph’s systematic testing of patient cells is the kind of transformative work the GigaLab was designed to support,” said Charlie Roco, PhD, co-founder and chief technology officer at Parse Biosciences. “This partnership shows how industrial-scale single cell biology and advanced AI can reveal disease mechanisms directly in patient cells and reshape drug discovery.”
