|
- Techniques for clustering gene expression data. - Genetic regulatory network (GRN) inferrence from microarray data with evolutionary computation - Genomic Analysis of Neuronal Remodelling
Techniques for clustering and pattern discovery in gene expression data PhD Student: Grainne KERR. Project description: Statistical methods can identify specific genes and groups of genes through co-expression but rather less about the pattern and dynamics of interaction. A natural representation of such inter-connectivity is a network. Traditionally such a network is considered as a (weighted) graph. For gene expression analysis, this can be organised as a bi-partite model or a one-mode model. Currently, Grainne is finalising her work on methods of graphical representation of information from gene expression datasets, and using tools and notions from classical network analysis to analyse these graphs. She is also focusing on cluster validation techniques, specifically for biclustered data. Genetic regulatory network (GRN) inferrence from microarray data with evolutionary computation PhD Student: Alina SIRBU. Project description: GRNs are networks of interactions between genes and the proteins that regulate their expression, responsible for different cell behaviour in different conditions. Uncovering GRNs is one of the most important issues in Systems Biology, as this would provide insights into the mecanisms involved in the functioning of organisms, or development of disease, and make way for the introduction of new and more efficient treatments. The evolution of high throughput technologies that measure gene expression levels have created a data base for inferring GRNs. However, the nature of this data has made this process very difficult. Several mathematical models and machine learning algorithms have been applied to microarray data but, while a large amount of biological information has been extracted, the task is far from being accomplished. This work attempts to develop a novel evolutionary computation approach to GRN inference. The aim is to integrate, along with microarray data, other types of biological data sets, in order to enhance the discovery of parameters for a finely-grained GRN model . The method will avail itself of parallel computation in order to make the approach scalable to larger scale networks. Genomic Analysis of Neuronal Remodelling Post-doctoral researcher: Dr. Laurence O'Dwyer Project Description: Axon pruning in the brain plays an important role in modifying neural connectivity in early development. Branches and connections in the immature nervous system are often pruned in order to ensure the proper formation of functional circuitry. Using microarray data from Drosophila experiments we are studying networks of genes that are involved in the localised degeneration of axons. The steroid hormone ecdysone initiates this degeneration by a reange of mechanisms including the up-regulation of genes involved in the ubiquitin-proteasome pathway. Proteolysis by the ubiquitin-proteasome pathway is also responsible for regulating vital functions of the nervous system such as synaptic plasticity and is associated with Alzheimer's Disease. By analysing time-series microarray data in wild-type and mutant neurons in which the ecdysone receptor is blocked, we aim to elucidate the roles of networks of genes in neuronal remodelling.
|
