BIRS Workshop Lecture Videos

Banff International Research Station Logo

BIRS Workshop Lecture Videos

Simulating cancer systems biology with PhysiCell: customized simulators, model exploration, and machine learning Heiland, Randy

Description

Cutting-edge cancer treatments like immunotherapy and engineered microbes are examples of current and upcoming cell-based therapeutics. The success, failure, and side effects of these therapies critically depend upon multicellular cancer systems biology: the dynamical chemical and mechanical interactions between the engineered cells, tumor cells, and the microenvironment. Computational models can act as "virtual laboratories" for multicellular systems. The ideal laboratory would include cell and tissue biomechanics, biotransport of multiple chemical substrates including signaling factors, and many interacting cells. In this talk, we will introduce PhysiCell (\url{http://dx.doi.org/10.1371/journal.pcbi.1005991}), an open source agent-based platform for 3D multicellular systems biology. With this platform, desktop workstations can routinely simulate systems of ten or more cell-secreted chemical signals and tissue substrates, along with $10^5$ to $10^6$ individual cells that grow, divide, die, secrete chemical signals, move, exchange mechanical forces, and remodel their tissue microenvironment. After introducing PhysiCell, we will describe some recent projects that help expand its use. One project demonstrates how customized Jupyter notebook GUIs for PhysiCell models can be automatically generated and ported to nanoHUB (\url{https://nanohub.org}) where they can be run by anyone using just a Web browser. This makes it possible to 1) provide easy-to-run simulators to accompany models discussed in publications, and 2) let students submit demo simulators as part of class assignments and also be part of their academic portfolio. Another project describes two extreme-scale model explorations using the EMEWS framework on Cray supercomputers at Argonne National Lab. In our most recent collaboration, we explored a six-parameter therapeutic design space for cancer immunotherapy. This combined high-performance/high-throughput computing and active learning to optimize the immunotherapeutic design, characterize the topology (shape) of the design space, and automatically rank the importance of design parameters. We will close by pointing to some of the free online simulators and briefly discuss the future outlook for using high-throughput multicellular simulations to efficiently explore high-dimensional design spaces and accelerate discovery. Joint work with: Gary An, Nicholson Collier, Jonathan Ozik, and Paul Macklin.

Item Media

Item Citations and Data

Rights

Attribution-NonCommercial-NoDerivatives 4.0 International