Investigating inter-replicate differences in cancer wound healing assays using an agent-based model

Cellular migration, and thus motility, are important factors in a tumors ability to metastasize. Wound healing assays are a way of quantifying these properties in vitro, while mathematical modeling can be used to do so in silico. In silico models can be calibrated and validated on the collected migration data, and used to predict the effects of therapeutics on the migration of cancer cells. We focus on the murine cell lines TC-1 and mEER, both transformed using the oncogenes HPV16 E6, HPV16 E7 and hRAS. Wound healing assays were performed on these cell lines after irradiation with 0Gy, 2Gy, 8Gy and 10Gy to quantify the dose dependent effects of radiation on the motility of the cell lines. Herein we report on a calibrated agent-based model used to investigate how assumptions about the underlying distributions of migration speed data affects in silico experiments. Additionally, we discuss whether combining the replicates per experiment lends itself to more accurate predictions than using each data set individually.