How Graph Therapeutics is rethinking immunology drug discovery using live patient cells, and why daphni is backing its mission through our Blue fund.
Can Testing Drugs on Real Human Immune Cells Improve Drug Discovery?
What if we could better predict whether an immune drug will work before it ever reaches a patient?
It's one of the biggest challenges in modern medicine. Despite decades of scientific progress, developing effective treatments for immune and inflammatory diseases remains extraordinarily difficult. Many promising drugs perform well in early experiments, only to fail when tested in humans.
The problem? The models used to discover drugs don't always reflect the complexity of the human immune system.
That's what Graph Therapeutics is setting out to change.
Backed by daphni through our Blue fund, Graph is developing a drug discovery platform that works directly with living human immune cells, helping researchers understand how potential treatments behave in real human biology before clinical trials begin.
Why do so many immune drugs fail in clinical trials?
Immune and inflammatory diseases affect hundreds of millions of people worldwide. Conditions such as rheumatoid arthritis, lupus, and psoriasis occur when the immune system mistakenly attacks healthy tissues or becomes chronically overactive.
Developing treatments for these diseases is particularly challenging because the human immune system is incredibly complex.
Traditional drug discovery often relies on simplified laboratory models, isolated cell systems, or animal experiments to predict how a treatment might work.
While these methods provide valuable insights, they cannot always reproduce the dynamic interactions occurring inside the human body.
The consequences are significant. Approximately nine out of ten drugs entering clinical trials fail to reach patients, across therapeutic areas. One contributing factor is that promising results from early experiments don't always translate into clinical effectiveness.
For immune diseases, where biological responses can vary considerably between patients, this disconnect is especially important.
What if researchers could study drug responses in real human immune cells much earlier in the development process?
How does Graph Therapeutics use live patient cells to discover better drugs?
Graph Therapeutics is building a platform designed to understand immune biology through direct experimentation on living human cells.
Rather than relying exclusively on simplified models, Graph studies how immune cells from real patient samples respond to different drugs and biological interventions.
At the heart of its technology is the Functional Immune Manifold (FIM), a system designed to map the relationships between immune dysfunction, disease states, and drug responses.
Graph combines several scientific approaches:
- Live human immune cells: Studying drug responses in patient-derived biological samples.
- High-content imaging: Capturing detailed information about how individual cells behave and change.
- Single-cell RNA sequencing: Examining gene expression at the level of individual cells to understand their biological responses.
- Lab-in-the-loop experimentation: Connecting experimental results with computational analysis to continuously refine the understanding of immune biology.
Together, these methods help Graph build a more detailed, functional picture of how the human immune system responds to potential treatments.
The goal is to identify promising drug targets and evaluate therapeutic candidates using evidence grounded in human biology, long before clinical trials begin.
What makes Graph Therapeutics' drug discovery platform different?
One of Graph's distinguishing features is its ability to study immune responses across different patients, disease backgrounds, and experimental conditions.
Instead of treating the immune system as a static collection of biological measurements, Graph is mapping how it actually behaves.
And the platform is already operational.
According to Graph's investment materials, the company has mapped approximately 1,000 disease-perturbation states across five disease areas.
This provides a foundation for identifying patterns of immune dysfunction and understanding how different interventions influence those patterns.
For pharmaceutical research, this approach could help answer several important questions earlier: Which biological targets are most relevant to a disease? How do different patient samples respond to a potential treatment? And which drug candidates deserve further development?
By bringing human biological evidence into earlier stages of discovery, Graph aims to improve the decisions that determine which treatments advance toward clinical testing.
