Integrative Analysis of BMI and Gene Expression Reveals Molecular Interactions Underlying Cancer Progression

Area of Science:

• Genomics
• Cancer Biology
• Bioinformatics

Background:

• Obesity, a chronic condition, is linked to various health issues, including increased cancer risk.
• High or low body mass index (BMI) can affect cancer risk and treatment outcomes.
• The
• obesity paradox
• suggests higher BMI may sometimes offer protective effects in cancer.

Purpose of the Study:

• To explore the relationship between BMI and gene expression in cancer using The Cancer Genome Atlas (TCGA) data.
• To identify potential links between BMI and cancer progression.
• To uncover potential therapeutic targets for obesity-related cancers.

Main Methods:

• A two-stage overlapping group screening (OGS) method was employed.
• Gene groups and interactions were identified using the "grpregOverlap" R package and sequence kernel association test.
• Predictive models were built using regularized regression techniques, including generalized ridge regression, lasso, and adaptive lasso.

Main Results:

• The OGS-based method, particularly OGS_G.ridge_ALasso, demonstrated superior prediction performance and stability over other models.
• Key BMI-associated genes and gene-gene interactions were identified in TCGA cancer patient data (bladder, cervical, esophageal, liver cancers).
• Network structures illustrating these interactions were presented.

Conclusions:

• The proposed method effectively predicts BMI-associated gene expression patterns in cancer.
• Identified BMI-associated genes and interactions serve as potential biomarkers for cancer development and prognosis.
• The study provides deeper insights into the biological mechanisms linking BMI and cancer.