A computational framework for predicting obesity risk based on optimizing and integrating genetic risk score and gene...

Area of Science:

• Genetics
• Obesity Research
• Computational Biology

Background:

• Large-scale genome-wide association studies (GWAS) have identified numerous genetic loci linked to Body Mass Index (BMI).
• Gene expression profiles also show associations with BMI, but accurate obesity risk prediction using genetic data remains difficult.
• Integrating diverse genetic factors is crucial for advancing obesity prediction models.

Purpose of the Study:

• To develop and evaluate a novel mathematical framework for predicting obesity genetic risk.
• To integrate Body Mass Index (BMI)-associated single nucleotide polymorphisms (SNPs) and gene expression profiles for enhanced prediction.
• To assess the performance of predictive models in a specific cohort.

Main Methods:

• A cohort of 75 individuals was analyzed, integrating 27 BMI-associated SNPs and obesity-associated gene expression data.
• A genetic risk score was calculated by summing BMI-increasing alleles, with an optimization algorithm used to refine SNP selection.
• Linear regression and support vector machine models were employed to predict obesity risk.

Main Results:

• An optimized genetic risk score showed significant correlation with BMI.
• Linear regression achieved an adjusted R-squared of 0.556.
• Support vector machine models demonstrated a prediction accuracy of 76% for obesity risk.

Conclusions:

• A new mathematical method for predicting obesity genetic risk was successfully developed and applied to a small cohort.
• The integrated computational framework combining genetic risk scores and gene expression shows promise for personalized obesity risk assessment.
• This approach provides a foundation for predicting obesity risk in specific populations using genetic information.