A systematic approach for developing mechanistic models for realistic simulation of cancer cell motion and deformation

Understanding and predicting metastatic progression and developing novel diagnostic methods can highly benefit from accurate models of the deformability of cancer cells. Spring-based network models of cells can provide a versatile way of integrating deforming cancer cells with other physical and biochemical phenomena, but these models have parameters that need to be accurately identified. In this study we established a systematic method for identifying parameters of spring-network models of cancer cells. We developed a genetic algorithm and coupled it to the fluid-solid interaction model of the cell, immersed in blood plasma or other fluids, to minimize the difference between numerical and experimental data of cell motion and deformation. We used the method to create a validated model for the human lung cancer cell line (H1975), employing existing experimental data of its deformation in a narrow microchannel constriction considering cell-wall friction. Furthermore, using this validated model with accurately identified parameters, we studied the details of motion and deformation of the cancer cell in the microchannel constriction and the effects of flow rates on them. We found that ignoring the viscosity of the cell membrane and the friction between the cell and wall can introduce remarkable errors.

Automated summary

This study established a systematic method for identifying parameters of cancer cell deformability models through the use of spring network models and genetic algorithms. By accurately simulating the deformation of human lung cancer cells in narrow microchannel constriction and analyzing the effects of flow rates on cell motion and deformation, it was found that neglecting cell membrane viscosity and cell-wall friction can introduce significant errors.