A graphical interface for running CUDA-powered viral infection simulations, built so researchers and students can use it without writing code.
Graduate students at TCU use a CUDA C agent-based model to study virus-mediated cell interactions. Previously, running the model meant editing parameter files by hand and launching simulations from the terminal, making it difficult for undergraduates and non-experts to use.
We built a PySide6 desktop GUI that provides intuitive controls for configuring simulation parameters, launching GPU-accelerated simulations, and visualizing the results. No command-line or code editing required.
The interface makes the model available for undergraduate research and classroom instruction. It also provides a well-documented, maintainable codebase that future students can extend with new biological processes.
Adjust cell counts, infection rates, transmission probabilities, and timing through structured input fields with real-time validation.
Choose between the Original viral transmission model (cell-to-cell and cell-free infection) or the Gerg syncytia formation model, with the UI adapting to show the relevant parameters.
Launch, monitor, and stop GPU-accelerated CUDA simulations from the interface. Progress feedback and automatic result loading are built in.
Plot uninfected, eclipse, infectious, and dead cell populations alongside virus concentration over time using interactive time-series graphs.
Save configurations as JSON, export results to CSV, and generate plot images. Spatial data output is also supported for further analysis.
Built-in validation catches out-of-range values and invalid configurations before a simulation can be launched, with clear error messages.
Key project documents. Click any card to download.