“All cells must be able to detect and respond to changes in their environment. Receptors located on the cell surface monitor external conditions and activate signaling pathways that, in turn, relay information about the environment to appropriate locations within the cell. Both in appropriate and suppressed activation of signaling pathways underlies many diseases including cancer and diabetes. Therefore, establishing how these systems function is important for developing new therapeutic strategies.
Because signaling pathways represent complex biochemical reaction networks, it is often not possible to understand their behavior without the aid of computational approaches. A main goal of our lab is to develop mathematical models that can predict the spatial and temporal dynamics of signaling pathways.
To accomplish this goal, we collaborate with experimental biologists who generate the data required to train and validate our models. We rely on the high-performance computing clusters maintained by Research Computing to simulate our models.
In particular, our particle-based simulations, which can involve thousands of molecules, require high-performance computing to run efficiently. Additionally, we use the clusters to run machine-learning algorithms to train our models on experimental data and perform model selection. These machine-learning approaches can involve running tens of thousands of simulations, again making high-performance computing a necessity. Finally, we rely on the Research Computing clusters to run computational image analysis tools required to extract the quantitative measurements needed for model training data sets.
In summary, the resources provided by Research Computing are allowing us to combine computational approaches with quantitative experimental investigations to gain new insights into the molecular mechanisms that regulate intracellular signaling pathways.”