Example PhD

Prototype Hybrid High-Throughput Virtual Cell Ensemble Modelling System

Supervisor: Professor N.J. Avis

Keywords: Cellular Modelling, Parallel Processing, GPUs, Visualization, Computational Steering

Large scale cellular modelling has provided important insights into myocardial function and pathophysiology and has now become a well-established component of the physiological armamentarium. To date, however, few studies have applied this approach to dynamic arterial activity - the phenomenon of vasomotion which serves as an adaptive mechanism to optimize microcirculatory perfusion and mass transfer processes under conditions of changing metabolic demand. Experimental and modelling studies of smooth muscle cells have provided insights into the importance of nonlinear control at the subcellular level, including the genesis of chaos. We wish to extend our previous modelling approach to realistic multicellular arterial architectures, including the important interaction between endothelial and smooth muscle cells.

Due to the computational demands of such work, we will harness the increasing range of multi-core computational platforms such as GPGPUs. The second element of the work is to provide support mechanisms to the scientists to provide user-friendly access to these devices and their outputs and hence promote HPC techniques into the biosciences. We will compare and contrast several computational approaches using the state-of-the-art facilities available within Cardiff and provide a prototype system capable of harnessing and presenting these resources to the biological scientist in an easy to use system. The system will support a workflow which enables the scientist to create a computational experiment, define its topology and the scope of the parameter sweep studies required. The system will then manage these tasks and run these on appropriate computational resources and report the results back to the scientist who will be able to visualize the spatial and temporal results and if required terminate particular parameter sweeps or refine existing ones.

Key Skills/Background: Open to Computing Graduates and Postgraduates

Contact: Professor N.J. Avis to discuss this research topic.