Modelling the Gene Dynamics of a Yeast Cell
The year 2003 marks the completion of the sequencing phase of human genome. Despite such remarkable and rapid progress in elucidating the structure and composition of eucaryotic genome, the problem of understanding the complexity of its function is largely unsolved. There is a large amount of gene expression data being produced in different labs with the intent of using it for genetic network inference. Typically there are two approaches that have been used so far in modelling genetic network. The qualitative and the quantitative approach. The qualitative approach is based on the notion of Boolean network as these networks share many properties such as complex behavior, self organization and periodicity with genetic system. In quantitative modelling, Baysian networks is one of the widely used approach. This project aims to combine both these approaches. The idea is to infer a Boolean network representing yeast's genetic network through an appropriate quantization of continuous microarray data representing gene expression profiles. As gene regulatory networks are mostly governed by Canalizing Boolean functions, we use this well known fact in reducing the set of Boolean networks consistent with the quantized data. Use of continuous representations of Boolean functions gives the set of Boolean networks in continuous version, all consistent with the quantized data. With the help of standard discrete optimization techniques, such as Least Square Optimization, the error between the set of modelled Boolean networks and the measured data set is optimized and an optimal network is obtained. This optimal network is then used for validation of the model. Weitere Informationen zu diesem Forschungsprojekt können Sie hier bekommenPublikationen
| |||||||||||||||