Multiscale Modeling of Biological Systems: Integrating Metabolism, ECM Dynamics, and
AI-Driven Workflows within the PhysiCell Framework
Multiscale models provide a powerful means of studying complex biological processes across multiple spatial and temporal scales. They capture both intracellular events, such as signaling and metabolism, and extracellular interactions where cells communicate and coordinate. By revealing how genetic or environmental deregulation affects diseas progression, these models can suggest strategies to restore healthy phenotypes. In this talk, I will present recent updates to the PhysiCell ecosystem developed at the Barcelona Supercomputing Center (BSC). I will introduce PhysiCelldFBA, an extension that
integrates dynamic flux balance analysis to simulate metabolism within individual agents, and PhysiMESS, a plugin designed to simulate the complex 3D architecture of the extracellular matrix fibers. To demonstrate the versatility of these tools, I will showcase several use cases, ranging from metabolic switching in E. coli to large-scale cancer growth models.
Finally, I will discuss a novel integration of mechanistic modeling and artificial intelligence: the development of Model Context Protocol (MCP) servers. These servers transform Large
Language Models (LLMs) into specialized modeling assistants, enabling them to leverage tools such as NeKO for Boolean network construction, maBoSS for stochastic simulations, and PhysiBoSS for multiscale modeling. This framework represents a significant step toward automating the design and execution of complex biological simulations.
Marco Ruscone is a postdoctoral researcher in the Life Science Department, coordinated by Professor Alfonso Valencia, at the Barcelona Supercomputing Center (BSC). He began his academic career in Turin, Italy, earning a bachelor’s degree in Physics and a master’s degree in Physics of Complex Systems for Biology from the Università degli Studi di Torino. He completed his PhD in Computational Biology at the Institut Curie in Paris, under the mentorship of Dr. Laurence Calzone, Dr. Andrei Zinovyev, and Dr. Vincent Noël. His research centers on developing multiscale models for cancer, employing a hybrid approach that combines agent-based modeling techniques, stochastic simulations on Boolean networks and AI automation.
Pour suivre la conférence en direct, cliquez ici.