Modeling non-genetic dynamics of cancer cell states measured by single-cell analysis: Uncovering bifurcations that explain why treatment either kills a cancer cell or makes it resistant

Abstract

Single-cell transcriptomics offers a new vista on non-genetic tumor cell plasticity, a neglected aspect of cancer. The gene expression state of each cell is governed by the gene regulatory network which represents a high-dimensional non-linear dynamical system that generates multiple stable attractor states and undergoes destabilizing bifurcations, manifest as critical transitions. Modeling clonal cell population as statistical ensembles of the same dynamical system, a index IC is derived for detecting destabilization towards critical transitions in single-cell molecular profiles. Therapy-induced bifurcation explains why treatment backfires: a drug-treated cell is imposed the binary choice to either apoptose or become a cancer-stem cell.