Skip to main content
Figure 10 | BioData Mining

Figure 10

From: Mining SOM expression portraits: feature selection and integrating concepts of molecular function

Figure 10

Overrepresentation maps of six selected gene sets containing between N set = 157 and 472 genes. Overrepresentation in each tile of the mosaic is calculated in units of log(pHG) using the hypergeometric distribution and color-coded (significance: maroon>red>yellow>green>blue). White areas indicate metagenes not containing genes from the respective set). Strongest overrepresentation of the different gene sets is found in different regions of the SOM (see red circles). Overrepresentation can be concentrated within one or a few adjacent metagenes (e.g. muscle contraction, panel c) or spread over different disjunct regions of the map (immune system, panel f).

Back to article page