Trainee Project
Title:
Simulation of the evolution of a muticellular tumor spheroid
Dates:
2025/03/03 - 2025/08/31
Student:
Supervisor(s):
Description:
Glioblastoma multiforme is the most frequent and aggressive of all brain tumors. Treatment is initially surgical, with as wide an excision as possible. Total excision is impossible, as the tumor infiltrates the healthy brain parenchyma. This leaves the infiltrating peripheral zone, which must be targeted by additional treatments such as radiotherapy and/or chemotherapy. Despite treatment, the prognosis for these tumors remains poor, not least because of the resistance mechanisms put in place by glioma tumor cells.
One of the research themes of CRAN's BioSiS department concerns the numerical modeling of glioblastoma progression in order to better understand its evolutionary dynamics and recurrence mechanisms. In particular, it has been shown that malignant cells are able to co-opt microglial and astrocytic cells to divert them from their immune surveillance functions, induce immunosuppression and thus promote tumor progression. We therefore aim to understand and better describe the balance of power between tumor cells and non-tumor cells in the microenvironment.
The aim of this internship is to exploit a numerical tool (PhysiCell, https://physicell.org/) to simulate the progression (in the sense of cell proliferation and invasion) of a spheroid composed of tumor cells and cells from the microenvironment. A spheroid is a 3D in vitro tumor model (in the biological sense of the term) which, unlike 2D models, mimics the multicellular organization and cellular heterogeneity found in tumor microregions in vivo. The simulator will be calibrated to reproduce the results of biological experiments carried out in the laboratory using spheroids composed of 3 cell types (U87, microglia and astrocyte). To do this, it will be necessary to be able to estimate certain model parameters from simulated or real data acquired from confocal microscopy images.
The work requested covers the following points:
- Getting to grips with the PhysiCell framework to simulate tumor progression in different populations of a spheroid.
- Bibliographical research on inverse problem-solving methods based on Bayesian inference, with or without PhusiCell.
- Selection of data to be collected from images and other measurements to be used as input for inference.
- Exploration of model validity after defining objective criteria for model comparison.
One of the research themes of CRAN's BioSiS department concerns the numerical modeling of glioblastoma progression in order to better understand its evolutionary dynamics and recurrence mechanisms. In particular, it has been shown that malignant cells are able to co-opt microglial and astrocytic cells to divert them from their immune surveillance functions, induce immunosuppression and thus promote tumor progression. We therefore aim to understand and better describe the balance of power between tumor cells and non-tumor cells in the microenvironment.
The aim of this internship is to exploit a numerical tool (PhysiCell, https://physicell.org/) to simulate the progression (in the sense of cell proliferation and invasion) of a spheroid composed of tumor cells and cells from the microenvironment. A spheroid is a 3D in vitro tumor model (in the biological sense of the term) which, unlike 2D models, mimics the multicellular organization and cellular heterogeneity found in tumor microregions in vivo. The simulator will be calibrated to reproduce the results of biological experiments carried out in the laboratory using spheroids composed of 3 cell types (U87, microglia and astrocyte). To do this, it will be necessary to be able to estimate certain model parameters from simulated or real data acquired from confocal microscopy images.
The work requested covers the following points:
- Getting to grips with the PhysiCell framework to simulate tumor progression in different populations of a spheroid.
- Bibliographical research on inverse problem-solving methods based on Bayesian inference, with or without PhusiCell.
- Selection of data to be collected from images and other measurements to be used as input for inference.
- Exploration of model validity after defining objective criteria for model comparison.
Keywords:
Computer simulation, Inverse problem, Bayesian inference, Image processing, Oncology
Department(s):
| Biology, Signals and Systems in Cancer and Neuroscience |