By Morgan Nwanguma
Scientists at the University of Maryland School of Medicine (UMSOM) have created software that uses genomic data and mathematical modelling to forecast how cancer cells behave—much like predicting the weather. By pairing patient-specific genetic information with a novel plain-language “hypothesis grammar,” the tool simulates how cells communicate and evolve within tissues.
These digital forecasts allow researchers to explore, in silico, how cancers grow, how immune systems respond, and how different treatments might affect an individual patient. Just as meteorologists track storms, the software predicts shifts in cell activity over time, including the signalling that enables cancer to spread.
The work, co-led by UMSOM’s Institute for Genome Sciences (IGS), was published July 25 in Cell. It represents the culmination of a multi-year, multi-laboratory effort uniting computational, bench, and clinical scientists. Ultimately, the approach could pave the way for “digital twin” of patients—computer models that help doctors identify the most effective treatment strategies for each case of cancer.
“Although standard biomedical research has made immeasurable strides in characterizing cellular ecosystems with genomics technologies, the result is still a single snapshot in time — rather than showing how diseases, like cancer, can arise from communication between the cells,” said Jeanette Johnson, PhD, a Postdoc Fellow at the Institute for Genome Sciences (IGS) at UMSOM and co-first author of this study. “Cancer is controlled or enabled by the immune system, which is highly individualized; this complexity makes it difficult to make predictions from human cancer data to a specific patient.”
What sets this research apart is its use of a plain-language “hypothesis grammar,” which acts as a bridge between biology and computation by translating common language into cell behaviour models.
Developed by a team led by Paul Macklin, PhD, Professor of Intelligent Systems Engineering at Indiana University, the grammar allows scientists to describe how cells behave using simple English sentences. These descriptions can then be converted into digital simulations of multicellular systems, enabling researchers to model complex diseases such as cancer with unprecedented clarity.
“As much as this new ‘grammar’ enables communication between biology and code, it also enables communication between scientists from different disciplines to leverage this modelling paradigm in their research,” said Daniel Bergman, PhD, a scientist at IGS and Assistant Professor of Pharmacology and Physiology at UMSOM and co-leading author with Dr. Johnson.
Dr. Bergman and his colleagues at IGS combined the plain-language grammar with genomic data from patient samples—leveraging advanced tools such as spatial transcriptomics—to investigate breast and pancreatic cancers.
In breast cancer, their models revealed how the immune system can sometimes fail to restrain tumour growth and instead drive invasion and metastasis. They then adapted this computational framework to simulate an actual pancreatic cancer immunotherapy trial.
By integrating genomic data from untreated pancreatic tumour samples, the model predicted highly variable responses among virtual “patients,” underscoring the role of the tumour microenvironment in precision oncology. Pancreatic cancer is notoriously difficult to treat in part because tumours are surrounded by dense layers of fibroblasts—non-cancerous cells that shape how cancer progresses. Using spatial genomics, the team showed in detail how fibroblasts signal to tumour cells, and how these interactions fuel tumour invasion. The software made it possible to trace the progression of pancreatic tumours directly from patient-derived tissue.
“What makes these models so exciting to me as someone who studies immunology is that they can be informed, initialized, and built upon using both laboratory and human genomics data,” said Dr. Johnson. “Immune cells are amazing and follow rules of behaviour that can be programmed into one of these models. So, for instance, we can take data and treat it as a snapshot of what the human immune system is doing, and this framework gives us a sandbox to freely investigate our hypotheses of what’s happening there over time without extra costs or risk to patients.”
“Ever since my transitioning from my training in weather prediction at the University of Maryland, College Park into computation, I have believed that we could apply the same principles to work across biological systems to make predictive models in cancer. I am struck by how many rules of biology we don’t yet know,” said Elana J. Fertig, PhD, Director of IGS, Associate Director of Quantitative Sciences for the Greenebaum Comprehensive Centre, and Professor of Medicine and Epidemiology at UMSOM and a lead author on the study. “Adapting this approach to genomics technologies gives us a virtual cell laboratory in which we can conduct experiments to test the implications of cellular rules entirely in silico.”
Dr. Fertig described the work as “a tapestry of team science,” noting that clinical collaborators at Johns Hopkins University and Oregon Health & Science University helped validate the computational models. The project was supported by the National Foundation for Cancer Research.
Importantly, the new grammar is open source, ensuring that scientists everywhere can build upon it. “By making this tool accessible to the scientific community, we are providing a path forward to standardize such models and make them broadly adopted,” said Dr. Bergman.
To highlight its versatility, Genevieve Stein-O’Brien, PhD, the Terkowitz Family Rising Professor of Neuroscience and Neurology at Johns Hopkins School of Medicine, led a team that applied the approach to neuroscience. In this example, the programme successfully simulated how layered structures form during brain development, demonstrating the grammar’s potential well beyond cancer research.
“With this work from IGS, we have a new framework for biological research since researchers can now create computerized simulations of their bench experiments and clinical trials and even start predicting the effects of therapies on patients,” said Mark T. Gladwin, MD, Vice President for Medical Affairs at the University of Maryland, Baltimore, and the John Z. and Akiko K. Bowers Distinguished Professor and UMSOM Dean. “This has important applications to enable digital twins and virtual clinical trials in cancer and beyond. We look forward to future work extending this computational modelling of cancer to the clinic.”
Senior researchers on this study team are, Paul Macklin, PhD, Associate Dean for Undergraduate Education and Professor of Intelligent Systems Engineering at the Indiana School of Informatics, Computing and Engineering at Indiana University, Genevieve Stein-O’Brien, Bloomberg Assistant Professor of Neuroscience and Assistant Director Single-Cell Training and Analysis Centre (STAC) at Johns Hopkins University, and Dr. Fertig are continuing efforts to disseminate this software and extend its integration with genomics data for automatic model formulation through the National Cancer Institute (NCI) Informatics Technology in Cancer Research Consortium, who funded this study. Additional benchmarking of this study and applications of the software to breast and pancreatic cancer are supported from numerous NCI grants, the Jayne Koskinas Ted Giovanis Foundation, the National Foundation for Cancer Research, the Cigarette Restitution Fund Programme from the State of Maryland, and the Lustgarten Foundation.
English