Figure 1

Statistical analysis of colon cancer methylome. Top-left displays CpG-level methylation measurements for colon normal and tumor tissue for a sample of TCGA [12] data; top-right displays expression data from the Gene Expression Barcode Project across multiple cancer types as a heatmap, the hierarchical clustering is dynamically updated as the users navigates across the genome; the bottom display shows smooth function $g_k\left(l\right)$, described in the main text, corresponding to differences in methylation between normal and tumor; the "Colon Blocks" track displays statistically significant DMRs inferred from the smoothed methylation difference function. The brushing interaction (in yellow) links data from Bioconductor objects produced at different stages of a statistical analysis pipeline: from measurement preprocessing to statistically significant regions of interest allowing effective exploration of the statistical properties of these genomic findings. http://epiviz.cbcb.umd.edu/?ws=sp9ShCJdS3c