Scientists develop virtual cell lab to simulate biological processes and drug effects before live experiments

Scientists have made progress in developing a “virtual cell lab” capable of simulating biological processes and predicting cellular behavior, offering a cost-effective alternative to traditional experiments with live cells. The initiative, led by researchers from Indiana University, Johns Hopkins Medicine, the University of Maryland School of Medicine, and Oregon Health & Science University, aims to create a digital twin of cellular systems to test drugs, study gene-environment interactions, and analyze complex molecular dynamics in human biology. The project builds on PhysiCell, a computational tool originally developed by Paul Macklin, Ph.D., an engineering professor at Indiana University. PhysiCell uses “agents”—mathematical models representing cells—to simulate behaviors based on rules derived from DNA and RNA. These agents can interact with environmental factors like oxygen levels, therapeutics, and other molecules, enabling researchers to observe processes such as tumor formation, immune cell activity, and brain development. A key innovation is a new “grammar” for the software, designed to simplify its use for biologists without advanced programming skills. Stein-O'Brien, a neuroscience and neurology professor at Johns Hopkins, described the system as akin to an Excel spreadsheet, where each line links a cell type to a rule in human-readable language. For example, a rule might state: “This cell increases division as oxygen concentration rises.” The program automatically converts these rules into mathematical equations, generating predictive models of cell behavior. This approach reduces the time required to build simulations from months to hours, making it more accessible for researchers. The team has already applied the tool to study brain development and cancer. In one example, David Zhou, a Johns Hopkins neuroscience undergraduate student, and Zachary Nicholas, a human genetics Ph.D. candidate, created a model of brain development using data from the Allen Brain Atlas. The software uses spatially resolved data to reconstruct dynamic interactions between cells and tissues over time, a critical step for understanding human diseases. Another study focused on cancer cell behavior, leveraging data from human pancreatic tumors and mouse experiments. Jeanette Johnson, Ph.D., a postdoctoral fellow at the University of Maryland School of Medicine, validated the program by simulating how macrophages—a type of immune cell—invade breast tumors by activating the EGFR genetic pathway, which typically promotes cancer growth. The simulation predicted increased tumor mobility, a finding later confirmed with live cancer cells in the lab. Elana Fertig, Ph.D., a professor at the University of Maryland School of Medicine, co-led the project, emphasizing its potential to prioritize hypotheses and therapeutic targets before costly lab experiments. “We can focus our benchwork on what seems most promising,” she said. The researchers plan to expand the tool’s capabilities by integrating artificial intelligence to generate simulation models automatically, enabling faster adaptation to new data. This could enhance digital twin technologies for personalized medicine and disease research. While the program remains a work in progress, its development highlights the growing role of computational biology in advancing medical science. By bridging the gap between biological knowledge and digital modeling, the virtual cell lab could revolutionize how scientists study cellular systems and accelerate discoveries in areas like oncology, neuroscience, and immunology.