Unveiling Epigenetic Regulatory Elements Associated with Breast Cancer Development

Area of Science:

• Genomics
• Epigenetics
• Bioinformatics

Background:

• Breast cancer impacts millions globally, necessitating advanced research into its underlying mechanisms.
• Identifying reliable biomarkers and understanding gene expression regulation are critical for effective treatment strategies.

Purpose of the Study:

• To uncover epigenetic mechanisms influencing breast cancer gene expression.
• To discover novel breast cancer biomarkers.
• To develop an integrated bioinformatic approach combining feature selection, Natural Language Processing, and 3D chromatin analysis.

Main Methods:

• Utilized The Cancer Genome Atlas (TCGA) multi-omics data (mRNA, miRNA, DNA methylation) from over 800 samples.
• Applied Monte Carlo Feature Selection and Interdependency Discovery to reduce 417,486 features to 2701 significant ones.
• Integrated Natural Language Processing and 3D chromatin structure analysis.

Main Results:

• Achieved high classification accuracy between cancer and control samples using selected features.
• Observed generally lower differentially expressed gene (DEG) expression and increased differentially methylated site (DMS) β-values in cancer samples.
• Identified specific DMSs impacting transcription factor binding (NRF1, MXI1) and altering gene expression (NKAPL, PITX1).
• 3D chromatin models revealed looser packing in cancer cells.

Conclusions:

• The study highlights complex regulatory dependencies in breast cancer.
• The proposed bioinformatic approach effectively reduces data dimensionality and identifies key features.
• Findings provide a foundation for experimental validation of identified biomarkers and regulatory pathways.