Identification of markers associated with global changes in DNA methylation regulation in cancers

Abstract

DNA methylation exhibits different patterns in different cancers. DNA methylation rates at different genomic loci appear to be highly correlated in some samples but not in others. We call such phenomena conditional concordant relationships (CCRs). In this study, we explored DNA methylation patterns in 12 common cancers using data of 2434 patient samples collected by The Cancer Genome Atlas project. We developed an exploratory method to characterize CCRs in the methylation data and identified the 200 gene markers whose on-and-off statuses in DNA methylation are most significantly associated with drastic changes in CCRs throughout the genome. Clustering analysis of the methylation data of the 200 markers showed that they are tightly associated with cancer subtypes. We also generated a library of the significant CCRs that may be of interest to future studies of the regulation network of DNA methylation in cancer.

Identification of thresholds for dichotomizing DNA methylation data

“…For example, Figure 6 shows that the methylation of SOX8 is significantly associated to low SOX8 expression in breast cancer and kidney renal clear cell carcinoma but not in the other three cancer types. Such tissue-specific relationship echoes a previous result that hierarchical clustering of methylation data is able to separate tissue types and cancer subtypes [14,15]. Four genes showed significant methylation-expression association in all five individual cancer types, BST2, SLA2, GSTT1, and GSTM1.…”

Topological Characterization of Human and Mouse m5C Epitranscriptome Revealed by Bisulfite Sequencing

“…In DNA methylation, it has been shown that the correlation of methylation rates between two CpG sites is related to the distance (see Figure S4 ), and the clustering effect can be as high as 0.7 for probes within 200 bp [ 53 ]. To address whether the mRNA m 5 C methylation also exhibits a clustering effect, we examined the proportion of m 5 C sites that are within 10 bp distance of other m 5 C sites and compared this proportion with that from 1000 times of random permutation.…”

TreeVis: A MATLAB-based tool for tree visualization

“…The samples are composed of cell lines and 12 cancer types. We selected 200 CpG sites using conditional correlation [13], and used the methylation data of those CpG sites to build a minimum-spanning tree. The tree describes the similarities among the samples according to their methylation status of the selected CpG sites.…”