Development and validation of a novel cell type estimation method for targeted bisulfite sequencing data.

Aims: Epidemiological studies of DNA methylation often use buccal swabs, which contain mixtures of cell types, but no low-cost methods exist for statistical correction in candidate gene studies using targeted bisulfite sequencing. This study aims to address this gap by estimating the proportion of buccal epithelial cells in swab and mouthwash samples.

Materials & methods: We applied a recently described and smoothly to implement method for estimating the proportion of buccal epithelial cells in buccal swab and mouthwash samples using targeted bisulfite sequencing. Additionally, we investigated the methylation of actinin alpha 3 (ACTN3), a marker for cell type-specific methylation, following psychosocial and physical stress.

Results: The proposed estimation method showed strong correlation with the EpiDISH algorithm and effectively controlled for cellular heterogeneity. Over 90% of the variance in ACTN3 methylation was explained by including the epithelial cell proportion in the model.

Conclusion: Our findings provide a solution for controlling cellular heterogeneity in buccal swab and mouthwash DNA methylation studies. This method is particularly relevant for candidate gene studies in clinical settings, but future work should refine it for more detailed analyses of cell type proportions to improve precision.