A promising DNA methylation analysis pipeline for epigenetic studies and clinical implementation in inflammatory bowel disease

Background: Inflammatory bowel disease (IBD), a spectrum with two major conditions, Crohn’s disease and ulcerative colitis, is a growing burden for patients and health systems in developing countries. DNA methylation (DNAm) is the most studied epigenetic change and its correlation with IBD pathogenesis has been established. In this study, we developed a DNAm pipeline to process and analyze the DNAm data, which is potentially applied in IBD research and clinical practice. Methods: Our DNAm pipeline was based on the architecture of the Methylnet framework using deep learning to process and analyze DNAm data from multiple sequencing platforms and through the data analysis of both in-house and public datasets. Then, this model was validated against public datasets of age prediction. Results: We successfully developed a DNAm pipeline which is independent of sequencing platforms and easy to use. On validating against a public dataset, we confirmed that the DNAm pipeline is a great entry point to predict the phenotype of age with significant correlation with the ground truth (R2=0.96) and cell type deconvolution (highest R2=0.99 in neutrophils). Conclusions: DNAm, an epigenetic change in IBD, should be a target for investigation as it is linked to the disease. Although we did not have in-house IBD DNAm data yet, we expect that our DNAm data pipelin will create a foundation for finding biomarkers in IBD.