Dataset Information

CpGFilter: model-based CpG probe filtering with replicates for epigenome-wide association studies.

ABSTRACT:

Summary

The development of the Infinium HumanMethylation450 BeadChip enables epigenome-wide association studies at a reduced cost. One observation of the 450K data is that many CpG sites the beadchip interrogates have very large measurement errors. Including these noisy CpGs will decrease the statistical power of detecting relevant associations due to multiple testing correction. We propose to use intra-class correlation coefficient (ICC), which characterizes the relative contribution of the biological variability to the total variability, to filter CpGs when technical replicates are available. We estimate the ICC based on a linear mixed effects model by pooling all the samples instead of using the technical replicates only. An ultra-fast algorithm has been developed to address the computational complexity and CpG filtering can be completed in minutes on a desktop computer for a 450K data set of over 1000 samples. Our method is very flexible and can accommodate any replicate design. Simulations and a real data application demonstrate that our whole-sample ICC method performs better than replicate-sample ICC or variance-based method.

Availability and implementation

CpGFilter is implemented in R and publicly available under CRAN via the R package 'CpGFilter'.

Contact

chen.jun2@mayo.edu or xlin@hsph.harvard.edu

Supplementary information

Supplementary data are available at Bioinformatics online.

SUBMITTER: Chen J

PROVIDER: S-EPMC4757944 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundIn DNA methylation analyses like epigenome-wide association studies, effects in differentially methylated CpG sites are assessed. Two kinds of outcomes can be used for statistical analysis: Beta-values and M-values. M-values follow a normal distribution and help to detect differentially methylated CpG sites. As biological effect measures, differences of M-values are more or less meaningless. Beta-values are of more interest since they can be interpreted directly as differences in percentage of DNA methylation at a given CpG site, but they have poor statistical properties. Different frameworks are proposed for reporting estimands in DNA methylation analysis, relying on Beta-values, M-values, or both.ResultsWe present and discuss four possible approaches of achieving estimands in DNA methylation analysis. In addition, we present the usage of M-values or Beta-values in the context of bioinformatical pipelines, which often demand a predefined outcome. We show the dependencies between the differences in M-values to differences in Beta-values in two data simulations: a analysis with and without confounder effect. Without present confounder effects, M-values can be used for the statistical analysis and Beta-values statistics for the reporting. If confounder effects exist, we demonstrate the deviations and correct the effects by the intercept method. Finally, we demonstrate the theoretical problem on two large human genome-wide DNA methylation datasets to verify the results.ConclusionsThe usage of M-values in the analysis of DNA methylation data will produce effect estimates, which cannot be biologically interpreted. The parallel usage of Beta-value statistics ignores possible confounder effects and can therefore not be recommended. Hence, if the differences in Beta-values are the focus of the study, the intercept method is recommendable. Hyper- or hypomethylated CpG sites must then be carefully evaluated. If an exploratory analysis of possible CpG sites is the aim of the study, M-values can be used for inference.

Project description:BackgroundEpigenetic mechanisms have been suggested to play a role in the development of post-traumatic stress disorder (PTSD). Here, blood-derived DNA methylation data (HumanMethylation450 BeadChip) collected prior to and following combat exposure in three cohorts of male military members were analyzed to assess whether DNA methylation profiles are associated with the development of PTSD. A total of 123 PTSD cases and 143 trauma-exposed controls were included in the analyses. The Psychiatric Genomics Consortium (PGC) PTSD EWAS QC pipeline was used on all cohorts, and results were combined using a sample size weighted meta-analysis in a two-stage design. In stage one, we jointly analyzed data of two new cohorts (N = 126 and 78) for gene discovery, and sought to replicate significant findings in a third, previously published cohort (N = 62) to assess the robustness of our results. In stage 2, we aimed at maximizing power for gene discovery by combining all three cohorts in a meta-analysis.ResultsStage 1 analyses identified four CpG sites in which, conditional on pre-deployment DNA methylation, post-deployment DNA methylation was significantly associated with PTSD status after epigenome-wide adjustment for multiple comparisons. The most significant (intergenic) CpG cg05656210 (p = 1.0 × 10-08) was located on 5q31 and significantly replicated in the third cohort. In addition, 19 differentially methylated regions (DMRs) were identified, but failed replication. Stage 2 analyses identified three epigenome-wide significant CpGs, the intergenic CpG cg05656210 and two additional CpGs located in MAD1L1 (cg12169700) and HEXDC (cg20756026). Interestingly, cg12169700 had an underlying single nucleotide polymorphism (SNP) which was located within the same LD block as a recently identified PTSD-associated SNP in MAD1L1. Stage 2 analyses further identified 12 significant differential methylated regions (DMRs), 1 of which was located in MAD1L1 and 4 were situated in the human leukocyte antigen (HLA) region.ConclusionsThis study suggests that the development of combat-related PTSD is associated with distinct methylation patterns in several genomic positions and regions. Our most prominent findings suggest the involvement of the immune system through the HLA region and HEXDC, and MAD1L1 which was previously associated with PTSD.

Project description:BackgroundGrowing evidence shows that epigenetic modifications play a role in Alzheimer's disease (AD). We performed an epigenome-wide association study (EWAS) to evaluate the DNA methylation differences using postmortem superior temporal gyrus (STG) and inferior frontal gyrus (IFG) samples.ResultsSamples from 72 AD patients and 62 age-matched cognitively normal controls were assayed using Illumina© Infinium MethylationEPIC BeadChip. Five and 14 differentially methylated positions (DMPs) associated with pathology (i.e., Braak stage) with p value less than Bonferroni correction threshold of 6.79 × 10-8 in the STG and IFG were identified, respectively. These cytosine-phosphate-guanine (CpG) sites included promoter associated cg26263477 annotated to ABCA7 in the STG (p = 1.21 × 10-11), and cg14058329 annotated to the HOXA5/HOXA3/HOXA-AS3 gene cluster (p = 1.62 × 10-9) and cg09448088 (p = 3.95 × 10-9) annotated to MCF2L in the IFG. These genes were previously reported to harbor DMPs and/or differentially methylated regions (DMRs). Previously reported DMPs annotated to RMGA, GNG7, HOXA3, GPR56, SPG7, PCNT, RP11-961A15.1, MCF2L, RHBDF2, ANK1, PCNT, TPRG1, and RASGEF1C were replicated (p < 0.0001). One hundred twenty-one and 173 DMRs associated with pathology in the STG and IFG, respectively, were additionally identified. Of these, DMRs annotated to 30 unique genes were also identified as significant DMRs in the same brain region in a recent meta-analysis, while additional DMRs annotated to 12 genes were reported as DMRs in a different brain region or in a cross-cortex meta-analysis. The significant DMRs were enriched in promoters, CpG islands, and exons in the genome. Gene set enrichment analysis of DMPs and DMRs showed that gene sets involved in neuroinflammation (e.g., microglia differentiation), neurogenesis, and cognition were enriched (false discovery rate (FDR) < 0.05).ConclusionsTwenty-two DMPs and 30 DMRs associated with pathology were replicated, and novel DMPs and DMRs were discovered.