Medulloblastomas comprise a diverse group of malignant tumors of the cerebellum and are the most common pediatric brain cancers (Northcott et al., 2012). In 2011, Cho et al. analyzed transcriptomes of a collection of 189 primary medulloblastoma tumors. Unsupervised clustering of this dataset identified six distinct molecular subgroups of medulloblastoma, including one that is driven by the tumorigenic activation of the Hedgehog pathway (Cho et al., 2011). Here we considered only the samples of the Hedgehog subtype. As a “control”, we used samples of another subtype from this dataset where the relevant oncogenic process is not Hedgehog signaling but rather photoreceptor activation and GABRA5 up-regulation. We projected the samples from these two subtypes into the space of the 50 hallmarks by means of single sample GSEA (Barbie et al., 2009). Single sample GSEA (ssGSEA, see Methods) estimates the degree of enrichment of gene sets in individual samples. Then we estimated the degree of association between each hallmark’s ssGSEA profile and the Hedgehog vs. photoreceptor activation phenotypic distinction using the Information Coefficient (IC, see Methods). In fact, any standard method for
