Related Experiment Video
Updated: Mar 25, 2026
Sample Preparation to Bioinformatics Analysis of DNA Methylation: Association Strategy for Obesity and Related Trait Studies
Published on: May 6, 2022
Estimation of Cell-Type Composition Including T and B Cell Subtypes for Whole Blood Methylation Microarray Data
Lindsay L Waite1, Benjamin Weaver2, Kenneth Day2
1Section on Statistical Genetics, Department of Biostatistics, School of Public Health, University of Alabama at BirminghamBirmingham, AL, USA; HudsonAlpha Institute for BiotechnologyHuntsville, AL, USA.
This study introduces a novel two-stage model to accurately estimate detailed cell types and subtypes within whole blood DNA methylation samples. The method enhances epigenetic research by providing precise cell composition data for association studies and confounding adjustment.
Area of Science:
- Epigenetics
- Immunology
- Bioinformatics
Background:
- DNA methylation levels are significantly influenced by the cellular composition of biological samples.
- Leukocytes in whole blood, commonly used for DNA methylation studies, comprise diverse cell types and subtypes.
- Accurate estimation of cell-type proportions is crucial for interpreting DNA methylation data and avoiding confounding.
Purpose of the Study:
- To develop and validate a novel two-stage model for estimating proportions of major leukocyte cell types and T/B cell subtypes in whole blood.
- To improve upon existing methods by enabling estimation of specific T and B cell subtypes (naïve, memory, regulatory).
- To provide a tool for enhancing the analysis of DNA methylation data from platforms like Illumina HumanMethylation450k and WGBS.
Main Methods:
- A two-stage modeling approach was developed to deconvolve cell-type proportions from DNA methylation data.
- The model estimates proportions for six main leukocyte types: CD4+ T cells, CD8+ T cells, monocytes, B cells, granulocytes, and natural killer cells.
- The model further refines estimates for subtypes of T and B cells, including naïve, memory, and regulatory populations.
Main Results:
- The proposed model demonstrated reliable estimation of leukocyte cell types and subtypes using both real and simulated datasets.
- Validation confirmed the accuracy and robustness of the cell-type proportion estimates.
- The method successfully differentiates proportions of naïve, memory, and regulatory T and B cell subtypes.
Conclusions:
- The developed two-stage model accurately estimates detailed leukocyte cell composition from DNA methylation data.
- These precise cell-type estimates are valuable for association studies and controlling for confounding in epigenome-wide association studies (EWAS).
- The method is adaptable to various genome-wide methylation data platforms, including HumanMethylation450k arrays and whole genome bisulfite sequencing (WGBS).
