ProCell

ProCell is a modeling and simulation framework designed to investigate cell proliferation dynamics. Unlike traditional approaches, it explicitly accounts for the inherent stochasticity of cell division events.

The tool directly manipulates raw data from flow cytometry experiments (such as GFP signal histograms) and leverages computational intelligence—specifically Swarm Intelligence—to calibrate the proliferation models. ProCell has been successfully applied to complex in vivo scenarios, including the modeling of cell proliferation in human acute myeloid leukemia xenografts.

Stochastic Simulation CUDA Accelerated

How ProCell Works

ProCell bridges experimental biology and computational modeling by ingesting real-world flow cytometry data to simulate future cell population states over a maximum simulation time ($T$).

Experimental Inputs:

• Initial cell fluorescence histograms (e.g., GFP signal across the population).
• The number of distinct sub-populations and their initial proportions.
• The mean and standard deviation of the division time for each sub-population.
• A baseline fluorescence minimum threshold.

Simulation Output: ProCell produces a highly accurate predicted histogram of GFP fluorescence after time $T$, allowing researchers to validate hypotheses about division rates and population heterogeneity.

Software & Tools

We have developed multiple versions of ProCell to ensure accessibility for biologists and high performance for computational scientists.

• ProCell (Python/GUI): The core framework is easily installable via pip (pip install procell). It features a user-friendly Graphical User Interface designed to simplify modeling, calibration, and simulation without requiring coding expertise.
• cuProCell (GPU Version): A CUDA-accelerated implementation designed to handle massive populations and strongly reduce the computational effort of stochastic simulations, scaling effectively on modern GPUs.