The sound of one hand clapping – gene networks in cancer

Author: Benno Schwikowski, PhD, group leader of the Systems Biology lab at Institute Pasteur, France

Cancer has been called ‘a disease of the genome’, meaning that to understand—and combat—it, we have to look at the biological functions of genes. The human genome, including the complete inventory of all human genes, has been studied in great detail over the past 50 years, leading to much progress in healing, detecting, and preventing disease.

However, understanding complex diseases, such as ovarian cancer, requires looking beyond individual genes at the context of their function; how they interact with other genes. Think of how you generate sound by clapping your hands – it is hard to explain on the basis of your right hand or your left hand alone.

Gene networks

When you draw genes one by one and connect them with a line if they perform a function together, you get gene networks. Looking at different regions of a gene network, instead of individual genes one by one, provides an important new perspective on the molecular processes that drive cancer. For instance, ‘hot spots’ appear, in which not only single genes, but multiple, interacting genes turn up as interesting candidates in the search for biologically relevant abnormalities.

As these networks are large, we need computers to visualize and analyze them efficiently. Together with other laboratories, our group at Institute Pasteur develops statistical and computational tools that allow us to rapidly and accurately identify such hot spots in very large interaction networks, and to visualize them on the computer. In creating interactive visualizations of large networks, with related clinical and biological data, we contribute to the systematic analysis of the data from HERCULES ovarian cancer samples at the level of gene networks. To see an example, you can go to http://gravity.pasteur.fr/hercules

Team work

Generating genome-scale data, assembling the resulting large datasets into networks, and then visualizing, analyzing, and interpreting them requires a very diverse set of skills. In the photo below you see next to me, on the left, Pierre, a trained immunologist in our laboratory. Only a few years ago, it would have been highly unusual for someone like him to work alongside other researchers from computer science and mathematics. Today, this collaboration across different disciplines within and across different research sites, spearheaded by projects such as HERCULES, is seen as a key factor to maximize our chances in the fight against complex diseases such as ovarian cancer.