Description

The Human Methylation Atlas tracks display genome-wide DNA methylation profiles from deep whole-genome bisulfite sequencing (WGBS) of 39 primary human cell types sorted from 205 healthy tissue samples. This comprehensive resource enables fragment-level analysis across thousands of unique markers, providing a detailed reference for cell-type-specific methylation patterns.

The 205 samples from 39 cell type groups are organized into two data types:

• Merged tracks display the combined methylation signal across all biological replicates for a given cell type, providing one representative track per cell type.


• Replicate tracks display the methylation signal for each individual sample and are available for detailed analysis.

DNA methylation patterns are highly reproducible across individuals of the same cell type (>99.5% identical), reflecting the robustness of cell identity programs.

Display Conventions and Configuration

Signal tracks display methylation beta values on a 0-1 scale, where 0 indicates fully unmethylated CpGs and 1 indicates fully methylated CpGs. A value of -1 indicates missing data. For optimal comparison across cell types, set the vertical viewing range to 0-1 with auto-scale off.

Track Colors

Tracks are colored by tissue/cell type category as follows:

Methods

Sample Collection and Sequencing

Primary human cells were isolated from freshly dissociated adult healthy tissues using fluorescence-activated cell sorting (FACS), yielding high-purity preparations across major cell lineages. A total of 205 samples representing 77 primary cell types were collected from 137 consenting donors and merged into 39 cell type groups based on methylation similarity. Average sample purity exceeded 90% as determined by flow cytometry, gene expression, and DNA methylation analysis. Some cell types showed lower purity, including colon fibroblasts (78%), smooth muscle cells (82%), endothelial cells (86%), and adipocytes (87%).

Several cell types are absent from the atlas, typically due to limited availability of primary material. These include osteoblasts, cholangiocytes, cells of the adrenal gland, urethral epithelium, and haematopoietic stem cells. Subpopulations of interest, such as distinct neuronal or lymphocyte subtypes, were also not resolved separately.

Whole-genome bisulfite sequencing was performed using 150 bp paired-end reads at an average sequencing depth of 30× (minimum 6.62×). Libraries were prepared using the Accel-NGS Methyl-Seq DNA library preparation kit and sequenced on the Illumina NovaSeq 6000 platform.

Processing and Analysis

Reads were mapped to the human genome (hg38) using bwa-meth, deduplicated with Sambamba, and processed into per-CpG methylation calls. The genome was segmented into 7.1 million non-overlapping methylation blocks using a multi-channel dynamic programming algorithm that identifies regions of homogeneous methylation across samples.

Cell-type-specific differentially methylated regions were identified using a one-versus-all comparison approach. Regions uniquely unmethylated in specific cell types were found to be enriched for transcriptional enhancers and tissue-specific transcription factor binding motifs.

Data processing was performed using wgbstools, an open-source computational suite for DNA methylation sequencing data representation, visualization, and analysis.

Data Access

The raw data for these tracks can be explored interactively using the Table Browser or the Data Integrator. For automated analysis, the data may also be queried from our REST API.

The complete dataset, including all WGBS data files and processed methylation calls, is available from GEO accession GSE186458.

For questions regarding the data, please contact Prof. Tommy Kaplan at the Hebrew University of Jerusalem.

Credits

Data generation and analysis were performed at the Hebrew University of Jerusalem by the Dor, Kaplan, and Glaser laboratories and collaborators. Sample collection involved collaboration with Hadassah Medical Center, Oregon Health & Science University, Karolinska Institute, and University of Alberta.

References

Loyfer N, Magenheim J, Peretz A, Cann G, Bredno J, Klochendler A, Fox-Fisher I, Shabi-Porat S, Hecht M, Pelet T et al. A DNA methylation atlas of normal human cell types. Nature. 2023 Jan;613(7943):355-364. PMID: 36599988

Loyfer N, Rosenski J, Kaplan T. wgbstools: a computational suite for DNA methylation sequencing data analysis. Life Sci Alliance. 2026 Apr;9(4):e202503514. PMID: 41611450