Predicting CpG methylation levels by integrating Infinium HumanMethylation450 BeadChip array data.

Area of Science:

• Epigenetics
• Genomics
• Computational Biology

Background:

• The Infinium HumanMethylation450 BeadChip (450K) array is a widely used, affordable technique for DNA methylation analysis.
• While generating vast amounts of data, the 450K array covers only ~1.6% of human genome CpGs, limiting comprehensive epigenetic studies.
• Understanding DNA methylation is crucial for numerous biological processes and diseases.

Purpose of the Study:

• To develop and compare computational models for significantly expanding the coverage of the Illumina 450K array.
• To enhance the utility of existing DNA methylation datasets by increasing CpG coverage.

Main Methods:

• Developed and compared novel computational models to predict DNA methylation levels.
• Utilized whole genome bisulfite sequencing and Illumina 450K array data from human H1 embryonic stem cells for model training and validation.
• Assessed model performance through correlation analysis and cross-validation on independent datasets.

Main Results:

• The developed computational models demonstrated a significant ~11-fold expansion in coverage compared to the standard 450K array.
• Predicted and measured DNA methylation levels showed strong correlation in human embryonic stem cells.
• The proposed model exhibited superior prediction accuracy over existing methods and demonstrated generalizability across different cell types.

Conclusions:

• The developed computational approach effectively expands DNA methylation analysis coverage, offering a more comprehensive view of the methylome.
• This method significantly enhances the value of existing 450K array data, facilitating deeper insights into epigenetics.
• The proposed model's generalizability makes it a powerful tool for future epigenetic research.