Post Genome-Wide Gene-Environment Interaction Study Using Random Survival Forest: Insulin Resistance, Lifestyle Factors, and Colorectal Cancer Risk.
Su Yon JungJeanette C PappEric M SobelZuo-Feng ZhangPublished in: Cancer prevention research (Philadelphia, Pa.) (2019)
Molecular and genetic pathways of insulin resistance (IR) connecting colorectal cancer and obesity factors in postmenopausal women remain inconclusive. We examined the IR pathways on both genetic and phenotypic perspectives at the genome-wide level. We further constructed colorectal cancer risk profiles with the most predictive IR SNPs and lifestyle factors. In our earlier genome-wide association gene-environmental interaction study, we used data from a large cohort of postmenopausal women in the Women's Health Initiative Database for Genotypes and Phenotypes Study and identified 58 SNPs in relation to IR phenotypes. In this study, we evaluated the identified IR SNPs and selected 34 lifestyles for their association with colorectal cancer risk in a total of 11,078 women (including 736 women with colorectal cancer) using a 2-stage multimodal random survival forest analysis. In overall and subgroup (defined via body mass index, exercise, and dietary-fat intake) analyses, we identified 2 SNPs (LINC00460 rs1725459 and MTRR rs722025) and lifetime cumulative exposure to estrogen (oral contraceptive use) and cigarette smoking as the most common and strongest predictive markers for colorectal cancer risk across the analyses. The combinations of genetic and lifestyle factors had much greater impact on colorectal cancer risk than any individual risk factors, and a possible synergism existed to increase colorectal cancer risk in a gene-behavior dose-dependent manner. Our findings may inform research on the role of IR in the etiology of colorectal cancer and contribute to more accurate prediction of colorectal cancer risk, suggesting potential intervention strategies for women with specific genotypes and lifestyles to reduce their colorectal cancer risk.
Keyphrases
- genome wide
- postmenopausal women
- dna methylation
- insulin resistance
- copy number
- metabolic syndrome
- bone mineral density
- body mass index
- polycystic ovary syndrome
- physical activity
- genome wide association
- adipose tissue
- type diabetes
- weight loss
- cardiovascular disease
- emergency department
- randomized controlled trial
- public health
- clinical trial
- mental health
- machine learning
- mass spectrometry
- quality improvement
- social media
- skeletal muscle
- high resolution
- electronic health record
- gene expression
- free survival
- single molecule
- artificial intelligence
- wastewater treatment
- deep learning