Correcting for cell-type heterogeneity in DNA methylation: a comprehensive evaluation.
Ontology highlight
ABSTRACT:
SUBMITTER: Rahmani E
PROVIDER: S-EPMC5548185 | biostudies-literature | 2017 Feb
REPOSITORIES: biostudies-literature
Publications
Correcting for cell-type heterogeneity in DNA methylation: a comprehensive evaluation.
Rahmani Elior E Zaitlen Noah N Baran Yael Y Eng Celeste C Hu Donglei D Galanter Joshua J Oh Sam S Burchard Esteban G EG Eskin Eleazar E Zou James J Halperin Eran E
Nature methods 20170201 3
Similar Datasets
Project description:BackgroundMany different methods exist to adjust for variability in cell-type mixture proportions when analyzing DNA methylation studies. Here we present the result of an extensive simulation study, built on cell-separated DNA methylation profiles from Illumina Infinium 450K methylation data, to compare the performance of eight methods including the most commonly used approaches.ResultsWe designed a rich multi-layered simulation containing a set of probes with true associations with either binary or continuous phenotypes, confounding by cell type, variability in means and standard deviations for population parameters, additional variability at the level of an individual cell-type-specific sample, and variability in the mixture proportions across samples. Performance varied quite substantially across methods and simulations. In particular, the number of false positives was sometimes unrealistically high, indicating limited ability to discriminate the true signals from those appearing significant through confounding. Methods that filtered probes had consequently poor power. QQ plots of p values across all tested probes showed that adjustments did not always improve the distribution. The same methods were used to examine associations between smoking and methylation data from a case-control study of colorectal cancer, and we also explored the effect of cell-type adjustments on associations between rheumatoid arthritis cases and controls.ConclusionsWe recommend surrogate variable analysis for cell-type mixture adjustment since performance was stable under all our simulated scenarios.
| S-EPMC4855979 | biostudies-literature
Project description: Not available
| S-EPMC6659733 | biostudies-literature
Project description:Background/objectivesThe heterogeneity of the tumor immune microenvironment is a key determinant of tumor oncogenesis. This study aims to evaluate the composition of seven immune cells across 5323 samples from 14 cancers using DNA methylation data.MethodsA deconvolution algorithm was proposed to estimate the composition of seven immune cells using 1256 immune cell population-specific methylation genes. Based on the immune infiltration features of seven immune cell fractions, 42 subtypes of 14 tumors (2-5 subtypes per tumor) were identified.ResultsSignificant differences in immune cells between subtypes were revealed for each cancer. The study found that the methylation values of the selected specific sites correlated with gene expression in most tumor subtypes. Immune infiltration results were integrated with phenotypic data, including survival data and tumor stages, revealing significant correlations between immune infiltration and phenotypes in some tumors. Subtypes with high proportions of CD4+ T cells, CD8+ T cells, CD56+ NK cells, CD19+ B cells, CD14+ monocytes, neutrophils, and eosinophils were identified, with subtype counts of 9, 24, 22, 13, 19, 9, and 11, respectively. Additionally, 2412 differentially expressed genes between these subtypes and normal tissues were identified. Pathway enrichment analysis revealed that these genes were mainly enriched in pathways related to drug response and chemical carcinogens. Differences in ESTIMATE scores for subtypes of seven tumors and TIDE scores for eight tumors were also observed.ConclusionsThis study demonstrates the intra-tumor and inter-tumor immune heterogeneity of pan-cancer through DNA methylation analysis, providing assistance for tumor diagnosis.
| S-EPMC11855777 | biostudies-literature
Project description:BackgroundPlasma cell-free DNA (cfDNA) is derived from cellular death in various tissues. Investigating the tissue origin of cfDNA through cell type deconvolution, we can detect changes in tissue homeostasis that occur during disease progression or in response to treatment. Consequently, cfDNA has emerged as a valuable noninvasive biomarker for disease detection and treatment monitoring. Although there are many methylation-based methods for cfDNA cell type deconvolution, a comprehensive and systematic evaluation of these methods has yet to be conducted.ResultsIn this study, we benchmark five methods: MethAtlas, cfNOMe toolkit, CelFiE, CelFEER, and UXM. Utilizing deep whole-genome bisulfite sequencing data from 35 human cell types, we generate in silico cfDNA samples with ground truth cell type proportions to assess the deconvolution performance of the five methods under multiple scenarios. Our findings indicate that multiple factors, including reference marker selection, sequencing depth, and reference atlas completeness, jointly influence the deconvolution performance. Notably, an incomplete reference with missing markers or cell types leads to suboptimal results. We observe performance differences among methods under varying conditions, underscoring the importance of tailoring cfDNA deconvolution analyses. To increase the clinical relevance of our findings, we further evaluate each method's performance in potential clinical applications using real-world datasets.ConclusionsBased on the benchmark results, we propose general guidelines to choose the suitable methods based on sequencing depth of the cfDNA data and completeness of the reference atlas to maximize the performance of methylation-based cfDNA cell type deconvolution.
| S-EPMC11660681 | biostudies-literature
Project description:BackgroundDNA methylation in CpG context is fundamental to the epigenetic regulation of gene expression in higher eukaryotes. Changes in methylation patterns are implicated in many diseases, cellular differentiation, imprinting, and other biological processes. Techniques that enrich for biologically relevant genomic regions with high CpG content are desired, since, depending on the size of an organism's methylome, the depth of sequencing required to cover all CpGs can be prohibitively expensive. Currently, restriction enzyme-based reduced representation bisulfite sequencing and its modified protocols are widely used to study methylation differences. Recently, Agilent Technologies, Roche NimbleGen, and Illumina have ventured to both reduce sequencing costs and capture CpGs of known biological relevance by marketing in-solution custom-capture hybridization platforms. We aimed to evaluate the similarities and differences of these four methods considering each platform targets approximately 10-13% of the human methylome.ResultsOverall, the regions covered per platform were as expected: targeted capture-based methods covered > 95% of their designed regions, whereas the restriction enzyme-based method covered > 70% of the expected fragments. While the total number of CpG loci shared by all methods was low, ~ 24% of any platform, the methylation levels of CpGs covered by all platforms were concordant. Annotation of CpG loci with genomic features revealed roughly the same proportions of feature annotations across the four platforms. Targeted capture methods comprise similar types and coverage of annotations and, relative to the targeted methods, the restriction enzyme method covers fewer promoters (~ 9%), CpG shores (~ 8%) and unannotated loci (~ 11%).ConclusionsAlthough all methods are largely consistent in terms of covered CpG loci, the commercially available capture methods result in covering nearly all CpG sites in their target regions with few off-target loci and covering similar proportions of annotated CpG loci, the restriction-based enrichment results in more off-target and unannotated CpG loci. Quality of DNA is very important for restriction-based enrichment and starting material can be low. Conversely, quality of the starting material is less important for capture methods, and at least twice the amount of starting material is required. Pricing is marginally less for restriction-based enrichment, and the number of samples that can be prepared is not restricted to the number of capture reactions a kit supports. However, the advantage of capture libraries is the ability to custom design areas of interest. The choice of the technique would be decided by the number of samples, the quality and quantity of DNA available and the biological areas of interest since comparable data are obtained from all platforms.
| S-EPMC5970534 | biostudies-literature
Project description:Methylation patterns of circulating cell-free DNA (cfDNA) contain rich information about recent cell death events in the body. Here, we present an approach for unbiased determination of the tissue origins of cfDNA, using a reference methylation atlas of 25 human tissues and cell types. The method is validated using in silico simulations as well as in vitro mixes of DNA from different tissue sources at known proportions. We show that plasma cfDNA of healthy donors originates from white blood cells (55%), erythrocyte progenitors (30%), vascular endothelial cells (10%) and hepatocytes (1%). Deconvolution of cfDNA from patients reveals tissue contributions that agree with clinical findings in sepsis, islet transplantation, cancer of the colon, lung, breast and prostate, and cancer of unknown primary. We propose a procedure which can be easily adapted to study the cellular contributors to cfDNA in many settings, opening a broad window into healthy and pathologic human tissue dynamics.
| S-EPMC6265251 | biostudies-literature
Project description:DNA methylation is an important epigenetic mark that plays a vital role in gene expression and cell differentiation. The average DNA methylation level among a group of cells has been extensively documented. However, the cell-to-cell heterogeneity in DNA methylation, which reflects the differentiation of epigenetic status among cells, remains less investigated. Here we established a gold standard of the cell-to-cell heterogeneity in DNA methylation based on single-cell bisulfite sequencing (BS-seq) data. With that, we optimized a computational pipeline for estimating the heterogeneity in DNA methylation from bulk BS-seq data. We further built HeteroMeth, a database for searching, browsing, visualizing, and downloading the data for heterogeneity in DNA methylation for a total of 141 samples in humans, mice, Arabidopsis, and rice. Three genes are used as examples to illustrate the power of HeteroMeth in the identification of unique features in DNA methylation. The optimization of the computational strategy and the construction of the database in this study complement the recent experimental attempts on single-cell DNA methylomes and will facilitate the understanding of epigenetic mechanisms underlying cell differentiation and embryonic development. HeteroMeth is publicly available at http://qianlab.genetics.ac.cn/HeteroMeth.
| S-EPMC6203689 | biostudies-literature
Project description:BackgroundSmoking impacts DNA methylation, but data are lacking on smoking-related differential methylation by sex or dietary intake, recent smoking cessation (<1 year), persistence of differential methylation from in utero smoking exposure, and effects of environmental tobacco smoke (ETS).MethodsWe meta-analysed data from up to 15,014 adults across 5 cohorts with DNA methylation measured in blood using Illumina's EPIC array for current smoking (2560 exposed), quit < 1 year (500 exposed), in utero (286 exposed), and ETS exposure (676 exposed). We also evaluated the interaction of current smoking with sex or diet (fibre, folate, and vitamin C).FindingsUsing false discovery rate (FDR < 0.05), 65,857 CpGs were differentially methylated in relation to current smoking, 4025 with recent quitting, 594 with in utero exposure, and 6 with ETS. Most current smoking CpGs attenuated within a year of quitting. CpGs related to in utero exposure in adults were enriched for those previously observed in newborns. Differential methylation by current smoking at 4-71 CpGs may be modified by sex or dietary intake. Nearly half (35-50%) of differentially methylated CpGs on the 450 K array were associated with blood gene expression. Current smoking and in utero smoking CpGs implicated 3049 and 1067 druggable targets, including chemotherapy drugs.InterpretationMany smoking-related methylation sites were identified with Illumina's EPIC array. Most signals revert to levels observed in never smokers within a year of cessation. Many in utero smoking CpGs persist into adulthood. Smoking-related druggable targets may provide insights into cancer treatment response and shared mechanisms across smoking-related diseases.FundingIntramural Research Program of the National Institutes of Health, Norwegian Ministry of Health and Care Services and the Ministry of Education and Research, Chief Scientist Office of the Scottish Government Health Directorates and the Scottish Funding Council, Medical Research Council UK and the Wellcome Trust.
| S-EPMC10825325 | biostudies-literature
Project description:Saliva is a widely used sample in epigenetic research with children due to its non-invasive nature. Since DNA methylation (DNAm) profile is cell type (CT) specific, salivary DNAm associations with exposures may be influenced by CT compositions, which is highly variable in saliva as it contains immune and buccal epithelial cells (BEC). Reference-based CT deconvolution and statistically adjusting estimated CT in DNAm analyses have become an increasingly common practice. However, careful examinations of how different reference panels may affect DNAm results and alternative approaches (e.g., stratification) are lacking. To scrutinize the best analytical strategies on pediatric salivary DNAm, the current study used 529 salivary DNAm samples obtained from children (mean age = 7.26 years, SD = 0.26 years) in the Family Life Project. Our results showed higher estimated CT variability with child than adult reference panels and highlighted the impact of these estimated CT discrepancies on DNAm associations with social variables (socioeconomic status). Stratifying salivary DNAm samples by BEC proportions detected a larger number of significant associations with biological variables (sex) and tissue-specific effect with cotinine level, a tobacco smoke-exposure biomarker. We provide analytical recommendations for future epigenetic research involving pediatric saliva samples.
| S-EPMC11601407 | biostudies-literature
Project description:BackgroundIntra-sample cellular heterogeneity presents numerous challenges to the identification of biomarkers in large Epigenome-Wide Association Studies (EWAS). While a number of reference-based deconvolution algorithms have emerged, their potential remains underexplored and a comparative evaluation of these algorithms beyond tissues such as blood is still lacking.ResultsHere we present a novel framework for reference-based inference, which leverages cell-type specific DNAse Hypersensitive Site (DHS) information from the NIH Epigenomics Roadmap to construct an improved reference DNA methylation database. We show that this leads to a marginal but statistically significant improvement of cell-count estimates in whole blood as well as in mixtures involving epithelial cell-types. Using this framework we compare a widely used state-of-the-art reference-based algorithm (called constrained projection) to two non-constrained approaches including CIBERSORT and a method based on robust partial correlations. We conclude that the widely-used constrained projection technique may not always be optimal. Instead, we find that the method based on robust partial correlations is generally more robust across a range of different tissue types and for realistic noise levels. We call the combined algorithm which uses DHS data and robust partial correlations for inference, EpiDISH (Epigenetic Dissection of Intra-Sample Heterogeneity). Finally, we demonstrate the added value of EpiDISH in an EWAS of smoking.ConclusionsEstimating cell-type fractions and subsequent inference in EWAS may benefit from the use of non-constrained reference-based cell-type deconvolution methods.
| S-EPMC5307731 | biostudies-literature