AMMPER-2: A spatially explicit agent-based model of microbial radiobiology with redox dye simulation

Deep space is a place where life has never evolved, yet humans hope to travel there. The journey poses a unique threat: ionizing radiation. Unlike the gentle sunlight that warms the Earth, this radiation consists of high-energy particles that can slice through DNA, the instruction manual inside every cell. For astronauts on long missions to Mars or beyond, this damage is a primary health risk. To understand how this invisible force affects living things, scientists often use the budding yeast Saccharomyces cerevisiae. It is a simple, single-celled organism that grows quickly and shares many biological traits with human cells. However, a gap remains in our knowledge. We can measure how radiation damages a single cell in a lab, and we can measure how a whole population of yeast grows in a space experiment, but connecting the two has been difficult. The challenge is to understand how the microscopic injury to one cell translates into the macroscopic signals—like changes in growth speed or chemical color—that automated spacecraft sensors record.

To bridge this gap, a team of researchers has developed and refined a computer program called AMMPER-2. This tool acts as a virtual laboratory, simulating a population of yeast cells as they drift through space and encounter radiation. The program does not just guess at the outcome; it builds the scenario from the ground up. It treats every single yeast cell as an individual agent with its own location, health status, and ability to repair damage. When a high-energy particle, such as a proton, strikes the virtual culture, the model calculates exactly where the energy lands. It then simulates the immediate aftermath: the creation of reactive oxygen species, which are unstable molecules that act like shrapnel, damaging nearby cells, and the subsequent attempts by those cells to survive and divide. The goal is to see if these microscopic interactions can reproduce the large-scale data that real spacecraft, like the BioSentinel mission, actually collect.

The researchers recently updated their model to include a feature that mirrors a specific chemical test used in spaceflight. In many space biology experiments, the nutrient broth contains a dye called alamarBlue. This substance changes color based on the metabolic activity of the cells. Healthy, active yeast cells turn the blue dye pink, and eventually colorless, as they consume it. By tracking these color changes, scientists can gauge the health of the entire population without opening the container. AMMPER-2 now simulates this chemical process. It uses a mathematical description of how the dye reacts to the yeast's metabolism, allowing the computer to predict not just how many cells are alive, but how the color of the liquid should change over time. The team tested this by comparing their computer predictions against real data from experiments where yeast was exposed to different doses of proton and gamma radiation.

The results showed that the model works well, particularly for the wild-type yeast, which has a full set of DNA repair tools. The computer successfully predicted the rate at which the dye changed color across a range of radiation doses, matching the real-world measurements with a high degree of accuracy. The model also handled a mutant strain of yeast that lacks a specific DNA repair gene, which makes it more sensitive to radiation. While the model captured the general trend of the mutant's decline, it struggled to fully replicate a specific behavior seen in the real data: at high radiation doses, the real mutant cells seemed to settle into a slow, steady state of reduced activity rather than continuing to decline as the model predicted. This suggests that the current version of the model treats cell damage as a simple switch—either a cell is healthy or it is dead—whereas real cells might enter a more complex, intermediate state of injury that slows them down without killing them immediately.

Beyond the biological simulation, the researchers made significant improvements to the software itself to make it more useful for scientists and educators. They added a visual interface that allows users to run simulations without needing to write code, turning a complex technical process into a simple dashboard where one can select the radiation dose and cell type and watch the results unfold. They also optimized the underlying calculations to run much faster on a standard laptop. One of the most interesting findings from their testing concerned the behavior of the reactive oxygen species. In the original model, these damaging molecules were treated as permanent and stationary, lingering forever at the spot where they were created. The new model allows them to drift away and fade over time, mimicking real chemistry. Surprisingly, the researchers found that this added complexity made very little difference to the final outcome of the simulations. The damage patterns remained largely the same, suggesting that for the specific conditions they modeled, the simple assumption of permanent damage was sufficient to explain the results.

The study also explored how different types of radiation affect the yeast. While protons travel in distinct tracks, like bullets, gamma radiation spreads its energy more evenly, like a fog. The model showed that this difference in distribution matters greatly. Because the proton tracks are so narrow, they often miss the small group of cells entirely, leaving many unharmed even at high doses. In contrast, the uniform spread of gamma radiation ensures that every cell receives some damage. This distinction helps explain why the mutant yeast, which cannot repair damage well, suffered much more severe population losses under gamma radiation than under protons, even when the total amount of radiation energy was the same.

Ultimately, AMMPER-2 serves as a powerful tool for interpreting the data from space missions. It allows researchers to test hypotheses about how radiation damages life before they ever launch a satellite. By simulating the journey from a single particle strike to a population-level signal, the model helps scientists understand what the numbers on a screen truly mean. While the model is not perfect and still relies on simplifications, it provides a clear, testable link between the invisible violence of deep space and the visible growth of life. As humanity prepares for longer journeys beyond Earth, tools like this will be essential for ensuring that the biological systems supporting our future in space remain healthy and understood.