CORNEA : Drawing large co-expression networks for genes

Gene co-expression is of great importance to scientists studying biology and molecular biology. To make sense of expression datasets, we present a tool, CORNEA, to create and visualize large co-expression networks. These networks allows end users to explore gene co-expression relationsships for large datasets with many genes. CORNEA is implemented as a web service at Lotus Base and the tool is described in detail in this thesis. We are limited by three requirements: Ease of maintenance, performance and limited resources. We create a system with high performance by deriving efficient ways to evaluate levels of co-expression as well as calculate co-expression network layouts. Through several experiments on real-world datasets, we evaluate the performance of co-expression measures, and the performance and quality of network layout algorithms. In conclusion, we establish the most suitable algorithm for generating co-expression networks and provide a tool for generating and visualizing large co-expression networks.