Inter-individual variation of DNA methylation and its implications for large-scale epigenome mapping - PubMed (original) (raw)

Inter-individual variation of DNA methylation and its implications for large-scale epigenome mapping

Christoph Bock et al. Nucleic Acids Res. 2008 Jun.

Abstract

Genomic DNA methylation profiles exhibit substantial variation within the human population, with important functional implications for gene regulation. So far little is known about the characteristics and determinants of DNA methylation variation among healthy individuals. We performed bioinformatic analysis of high-resolution methylation profiles from multiple individuals, uncovering complex patterns of inter-individual variation that are strongly correlated with the local DNA sequence. CpG-rich regions exhibit low and relatively similar levels of DNA methylation in all individuals, but the sequential order of the (few) methylated among the (many) unmethylated CpGs differs randomly across individuals. In contrast, CpG-poor regions exhibit substantially elevated levels of inter-individual variation, but also significant conservation of specific DNA methylation patterns between unrelated individuals. This observation has important implications for experimental analysis of DNA methylation, e.g. in the context of epigenome projects. First, DNA methylation mapping at single-CpG resolution is expected to uncover informative DNA methylation patterns for the CpG-poor bulk of the human genome. Second, for CpG-rich regions it will be sufficient to measure average methylation levels rather than assaying every single CpG. We substantiate these conclusions by an in silico benchmarking study of six widely used methods for DNA methylation mapping. Based on our findings, we propose a cost-optimized two-track strategy for mammalian methylome projects.

PubMed Disclaimer

Figures

Figure 1.

Figure 1.

DNA methylation variation among healthy individuals (schematic figure). This figure displays artificial DNA methylation data for two amplicons with two unrelated samples/profiles each, which were designed to illustrate the effect of the three measures of inter-individual variation used in this study. The typical amplicon with high overall methylation (blue profiles, top) has a relatively high pairwise deviation between means (_v_3) and a pairwise deviation between high-resolution profiles (_v_1) that is substantially lower than the deviation between mean and high-resolution profile (_v_2), which is reflected in a substantial correlation between the rising and falling of the DNA methylation profile curves over the length of the amplicon. In contrast, the typical amplicon with low overall methylation (red profiles, bottom) has a low pairwise deviation between means (_v_3) and similar values for pairwise deviation between high-resolution profiles (_v_1) and deviation between mean and high-resolution profile (_v_2), indicating that the fluctuations in the profiles are not inter-individually conserved and presumably random.

Figure 2.

Figure 2.

Effect of average amplicon methylation (left) and overlap with bona fide CpG islands (right) on inter-individual variation of DNA methylation. This figure shows the means of the three measures of DNA methylation variation as bar plots. In the left panel, values are reported separately for the top-25% most unmethylated amplicons with an average amplicon methylation of <11.5% (this threshold is motivated in the Materials and Methods section) and for the remaining 75% of amplicons. In the right panel, distinction is made between amplicons that overlap with a bona fide CpG island (4) and those that do not. In both cases, error bars represent 95% confidence intervals under the assumption of normal distribution and the _P_-values in the legends are based on two-sample, two-sided, _t_-tests between the group means for each measure.

Figure 3.

Figure 3.

Benchmarking results for experimental mapping of DNA methylation. This figure displays the results of in silico benchmarking of different DNA methylation mapping methods for all amplicons. The _y_-axis shows _v_method values for all experimental methods included in this study (A1–F9, described in Table 1) and for seven negative controls, which are based on guessing rules rather than on experimental data (G1–G7, described in Table 1). The standard boxplot format is used (boxes show center quartiles, whiskers extend to the most extreme data point which is no more than 1.5 times the interquartile range from the box) and outliers are hidden.

Similar articles

Cited by

References

    1. Bird A. DNA methylation patterns and epigenetic memory. Genes Dev. 2002;16:6–21. - PubMed
    1. Weber M, Schübeler D. Genomic patterns of DNA methylation: targets and function of an epigenetic mark. Curr. Opin. Cell Biol. 2007;19:273–280. - PubMed
    1. Gardiner-Garden M, Frommer M. CpG islands in vertebrate genomes. J. Mol. Biol. 1987;196:261–282. - PubMed
    1. Bock C, Walter J, Paulsen M, Lengauer T. CpG island mapping by epigenome prediction. PLoS Comput. Biol. 2007;3:e110. - PMC - PubMed
    1. Reik W. Stability and flexibility of epigenetic gene regulation in mammalian development. Nature. 2007;447:425–432. - PubMed

Publication types

MeSH terms

LinkOut - more resources