This thesis investigates whether subtle DNA methylation alterations can be detected in histologically normal breast tissue and whether such changes reflect early epigenetic drift preceding overt malignant transformation. The analysis is conducted on three independent genome-wide methylation datasets generated using Illumina 450K and EPIC arrays. A unified and fully reproducible analytical pipeline is implemented, including dataset harmonization, exploratory intra- and inter-cohort analyses, and rigorous pre-processing. Technical artefacts are removed through literature-based probe filtering, probe-type bias diagnostics, and β-to-M value transformation to ensure variance stabilization and statistical robustness. Exploratory analyses focus on characterizing the Normal–Adjacent axis within each dataset. Global methylation distributions and low-dimensional embeddings reveal strong structural similarity between the two groups, indicating that early alterations are subtle and localized rather than global.