Inference of gene regulatory networks based on gene expression data

Proteins regulate all important processes in a cell of a living organism. For the synthesis of proteins, messengerRNAs (mRNAs) serve as a template. The genes on the DNA act as a matrix for these mRNAs. This process is called gene expression. Depending on the state of the cell, the individual genes are expressed in different concentrations.

Microarrays offer a good experimental possibility to measure the concentration of many mRNAs simultaneously. Researcher observe the development of gene expression patterns in a cell in time series experiments. We use these time series to find the regulation of genes and to develop a model to explain the observed patterns. This model is based on a system of piecewise linear differential equations, which parameters are estimated out of the data.

In most experiments the number of genes far exceeds the number of time points, which is a main problem in analysing gene expression data. Therefore known statistical methods cannot be applied directly or do not provide acceptable results. So, to estimate the parameters for our model, we have to modify existing methods and develop new ones, respectively.

Schematic illustration of a gene regulatory network. Genes can either inhibit or activate themselves or other genes (gene 1 inhibits gene 2 and activates itself). Often products of genes form complexes, which then influence the expression of other genes (gene 1 and gene 2 activate gene 3 interdependently). Research partners")?> Karin Schnetz and Röbbe Wünschiers (Institute of Genetics, University of Cologne) and Andreas Burkovski (Institute of Biochemistry) Contact:")?> Jutta Gebert (gebert@zpr.uni-koeln.de) and Nicole Radde (radde@zpr.uni-koeln.de)