Genomic Analysis Using Regularized Regression in High-Grade Serous Ovarian Cancer

Area of Science:

• Genomics
• Cancer Biology
• Epigenetics

Background:

• High-grade serous ovarian cancer (HGSOC) is a complex malignancy.
• Its development is linked to genetic alterations, epigenetic changes, and germline variations.
• The interplay between tumor methylation, germline genetics, and gene expression in HGSOC requires further elucidation.

Purpose of the Study:

• To investigate the influence of tumor methylation and germline genetic variation on gene expression in HGSOC.
• To identify specific genes and regulatory elements involved in these associations.
• To assess the utility of penalized regression methods for integrative genomic analysis in HGSOC.

Main Methods:

• Utilized Elastic Net (ENET) penalized regression across three independent datasets.
• Analyzed data from over 470 HGSOC patients.
• Adjusted for somatic copy number alterations in the models.

Main Results:

• Germline variation, alone or with methylation, did not significantly impact HGSOC gene expression.
• A significant association was found between regional methylation and the expression of five genes: WDPCP, KRT6C, BRCA2, EFCAB13, and ZNF283.
• CpGs in the ENET models for BRCA2 and ZNF283 were enriched in regulatory elements.

Conclusions:

• Tumor methylation is a key factor influencing gene expression in HGSOC.
• The findings highlight specific genes and regulatory regions potentially critical for HGSOC development.
• Regularized regression methods offer a valuable approach for integrative genomic studies in cancer research.