(193ab) A Parallel Framework for Systematic Development of Multiscale Models Bridging Subcellular Biochemistry to Cell Population Dynamics

Cellular evolution takes place at much longer time scales compared to the subcellular molecular events in signaling and gene transcription. Bridging these processes at distinct time scales is imperative for the understanding of the evolutionary mechanisms of cells. Nevertheless, existing simulation techniques are inadequate to link such multiscale phenomena in a single model. Here, we introduce a new framework for the systematic development of multiscale cell population models. In an innovative approach, the framework expands a single-cell biochemical network model into a cell population model. By applying message passing interface (MPI) parallelism, it launches parallel simulations on a single-cell model and then treats each parallel thread as an individual cell object in a population. The parallel simulation threads (cell objects) evolve through death and division based on the intracellular network dynamics and temporal evolution of the extracellular environment. The framework simulates cell death by terminating a parallel thread. It simulates cell division by creating a new thread (daughter cell) from an existing thread (mother cell). The MPI mediates cell-to-cell and cell-environment communications. We demonstrate these capabilities by developing two cell population models using the framework. In one model, we treat the cells independent so that their death and division are not affected by other cells in the population. In the second model, we incorporate cellular interdependency from a quorum-sensing communication. In this latter model, each cell can affect other cells by modifying a shared growth environment. Using the models, we evaluate the computational performance of the framework. Our analysis indicates high model scalability and computation speed gain while a little loss in the accuracy of the model predictions.