A computational simulator (ALISON) combining a finite element and an agent based model was created. It describes the concentration of relevant molecules throughout the virtual tissue and the behaviour of individual cells. ALISON is fully programmable and integrates cell-cell variability.
ALISON was experimentally validated in both cell lines and patient derived cultures. In particular, the ability of this tool to recapitulate response to platinum-based chemotherapy and taxane agents was assessed, together with the study of cell growth in absence of treatment.
Our results highlight the role of heterogeneity in HGSOC progression. Early stage disease models are accurately captured by homogeneous simulated cell populations, while heterogeneity is essential for the description of later stage disease. Our simulations also identified a rise in population heterogeneity potentially associated with the development of treatment resistance. While further evidence is required to confirm this result experimentally, it underscores the role of computational models in biomedical research.
We have also created a procedure for the calibration of digital twins from standard clinical information. While testing on a larger cohort is necessary, this procedure allows to infer treatment sensitivity from commonly available clinical information.
These results have been recently published as preprint [1] and are currently under revision. They have also been presented at the 2024 VPH conference. In addition, two software tools for the analysis of in vitro data (invasion and adhesion) have been completed and published [2], [3]. The code for the simulator and these tools is freely available on github (
https://github.com/MarilisaCortesi/(öffnet in neuem Fenster)).
[1] Cortesi M., Liu D., Powell, E., Barlow E, Warton K., Giordano E & Ford C.E. (2024). Development and validation of a computational tool to predict treatment outcomes in cells from High-Grade Serous Ovarian Cancer patients. bioRxiv 2024.10.02.616212;
[2] Cortesi, M., Liu, D., Powell, E., Barlow, E., Warton, K., & Ford, C. E. (2024). Accurate Identification of Cancer Cells in Complex Pre‐Clinical Models Using a Deep‐Learning Neural Network: A Transfection‐Free Approach. Advanced Biology, 2400034.
[3] Cortesi M, Li J., Liu D., Guo T., Dokos S., Warton K.& Ford C.E. (2024). A novel approach for the quantification of single-cell adhesion dynamics from microscopy images. bioRxiv 2024.10.08.616409.
