Dr Jean-Christophe Nebel is an Associate Professor in the School of Computer Science & Mathematics where he conducts research in pattern recognition and machine learning that is applied to Computer Vision and Computational Biology. In both areas, he has led projects, the success of which has relied on access to the University's High Performance Computer (HPC). In particular, this allowed the production of four high-quality journal publications that will be submitted to REF2021.
As a project example, he supervised a part-time PhD student (2011-2018), Jad Abbass, who investigated approaches to speed up and improve the accuracy of existing software conceived to predict a protein's three-dimensional (3D) structure. This knowledge results in important insights into molecular mechanisms involved in many diseases, identifying putative drug targets and facilitating drug design. While wet laboratory techniques are available, their high time, cost or/and experimental constraints have motivated the development of alternative computational methods. Typically, the prediction of a new protein's 3D structure requires around 28 days of computation using a standard PC. As a consequence, the experiments presented in [1], where performance of our two novel approaches are compared with those of the standard one using a significant and representative set of 70 proteins, would have required around 16 years of computation on a PC. Fortunately, access to the University's HPC reduced that time to a few weeks making that research possible. Note, an additional couple of papers were published based on experiments of similar complexity.
[1] Customised fragments libraries for protein structure prediction based on structural class annotations, J. Abbass and J.-C. Nebel, BMC Bioinformatics, 16:136, 2015
Dr Kerry Brown is a plant ecologist and conservation biologist (PhD). His primary research focusses on interactions between forest ecosystems and environmental change, particularly understanding responses of natural ecosystems to land-use and climate change. Another non-mutually exclusive interest focusses on understanding how forest degradation influences loss of functional diversity, and, in turn, ecosystem processes and availability of ecosystem services in temperate and tropical regions. Dr Brown's research also has a strong spatial modelling component, highlighting the synergistic effects of land-use and climate change on biodiversity, as well as expansion and contraction of future plant distributions.
As a project example, Dr Brown and his PhD students used the High Performance Computer (HPC) to model how climate and land cover change influenced patterns of plant diversity across the entire island of Madagascar (Brown et al. 2015). The HPC allowed the team to model the distribution of 2,186 plants. Use of the HPC decreased the processing time for this work from six months to two weeks.
Additionally, they also used the HPC to investigate future range shifts of vulnerable and endangered plants on Madagascar (Yesuf et al. submitted), as well as modelling deforestation and degradation on Madagascar (Yesuf et al. 2019). This is a vitally important topic, since species range shifts driven by climate and land use change will have a significant impact on diversity patterns across the tropics.
If you need help and advice on any of Kingston University's IT systems please contact the IT Service Desk.