“Computational crystal ball” helps predict cell behavior
A new study presents a computational modeling system that enables scientists to predict how cells behave and change over time using an intuitive, language-based framework.
Researchers from the University of Maryland and Johns Hopkins University developed a “cell behavior grammar” that allows users to describe cellular interactions in plain language and run virtual experiments on systems such as tumors. To test the model’s broader applicability, the team used the Allen Institute’s Mouse Brain Atlas to simulate cell layer formation during brain development. In pancreatic cancer simulations using patient genomic data, the model showed that immune signals drive tumor invasion by increasing cell movement rather than cell division, a finding later confirmed experimentally. “We can take these snapshots and put them into this virtual cell laboratory and basically hit ‘play’,” said Genevieve Stein-O’Brien, describing how the system enables dynamic modeling of cell behavior.
This work highlights the role of Seattle-based research institutions in supporting computational biology through widely used data resources and tools.