An agent-based model for simulating viral infections

As we have seen in the past year, new viruses can spread rapidly and cause outbreaks that need a quick response from researchers to develop or re-purpose treatments. While experiments and clinical studies form the basis of this response, the data generated by these studies can be further leveraged through the use of mathematical models. Properly calibrated and validated mathematical models can make predictions about scenarios that are difficult to test experimentally, but are also faster and cheaper when testing possible treatment regimens. We have developed a realistic agent-based model of viral infections that runs on graphical processing units (GPUs), so it runs fast enough to simulate typical in vitro viral studies in a few hours. We present here testing and validation of the model for influenza infections and show that it can be calibrated to simulate different viral infections.