Researchers from Johns Hopkins Medicine, Indiana University, the University of Maryland School of Medicine, and Oregon Health & Science University have developed a powerful new computational platform that simulates the behavior of living cells—offering a virtual testing ground for biological research, drug development, and disease modeling. The system, built on an upgraded version of the PhysiCell software, uses agent-based modeling to represent each cell as a “math robot” governed by rules derived from DNA, RNA, and environmental factors.
This virtual cell lab allows scientists to observe how cells interact, form tissues, respond to therapies, and evolve over time—without the need for costly or invasive live-cell experiments. It’s designed to function as a “digital twin,” capable of predicting how real cells might behave under different conditions, such as cancer treatment or brain development. The platform has already been used to simulate tumor growth, immune cell invasion, and cortical layering in the brain, with results validated against lab-grown cells and patient-derived data.
One of the most transformative aspects of the system is its accessibility. Instead of requiring complex coding, researchers can input cell behavior rules using plain English in a spreadsheet—for example, “this cell increases division as oxygen concentration increases.” The software then converts these statements into mathematical equations that drive the simulation. This “hypothesis grammar” democratizes modeling, allowing biologists without programming expertise to build and test virtual experiments.
The team also integrated spatial transcriptomics data to visualize how cells function in 3D tissue environments. In one study, they modeled how macrophages infiltrate breast tumors by activating the EGFR pathway, a known driver of cancer spread. The simulation accurately predicted increased tumor mobility, which was later confirmed in lab experiments.
The platform is being expanded to simulate brain circuits, immune responses, and personalized cancer therapies. It’s also being trained with AI to automatically generate models from genomic data, accelerating hypothesis testing and therapeutic discovery. The researchers envision a future where digital cell labs guide real-world experiments, helping prioritize the most promising treatments and reducing trial-and-error at the bench.
