Genome-wide methylation studies have significantly advanced our understanding of colorectal adenocarcinoma progression and biomarker discovery. Aberrant DNA methylation plays a crucial role in gene expression regulation during cancer transformation, highlighting the need to identify differentially methylated regions (DMRs) as potential diagnostic and therapeutic markers. However, an integrated resource to explore and validate methylation alterations across colorectal cancer stages has been lacking. We aimed to develop a platform that integrates existing methylation data, systematically identifies DMRs and provides a tool for further investigation. We created a database combining Illumina HumanMethylation450K and EPIC data from normal colon, adenoma and adenocarcinoma tissues, comprising 2346 samples from 19 datasets. Methylation levels were analysed in six gene regions, and comparisons between tissue types were made using Mann-Whitney, Kruskal-Wallis and ROC tests. Both adenoma and adenocarcinoma samples exhibited a general decrease in methylation compared to healthy tissue. Differential methylation in genes such as ITGA4, NPY, IGFL1 and LRRC4 was validated. The strongest DMRs were observed in the C1orf70 gene's 5'UTR and TSS200 regions, with AUC values of 0.98 in both of the HM450K and EPIC datasets. We established an interactive web-based platform accessible at https://epigenplot.com/ enabling future analysis of individual gene regions. Our study provides an integrated database of DNA methylation profiles across normal, adenoma and adenocarcinoma tissues, offering a valuable resource for biomarker discovery. The integrated web platform can serve as a tool for the development of methylation-based therapies in the future.
Read full abstract