Logistic Bayesian LASSO for detecting association combining family and case-control data.

Area of Science:

• Genetics
• Statistical genetics
• Bioinformatics

Background:

• Limited statistical power when analyzing case-control or trio data separately for genetic studies.
• Need for methods that integrate diverse family-based and case-control genetic data.
• Complex diseases often involve rare genetic variants requiring powerful detection methods.

Purpose of the Study:

• To propose eLBL, an extension of Logistic Bayesian LASSO, for joint analysis of case-control and trio data.
• To enhance statistical power for detecting associations between rare haplotypes and complex diseases.
• To account for familial correlations within the analyzed samples.

Main Methods:

• Developed eLBL, integrating Logistic Bayesian LASSO with joint analysis of case-control and trio data.
• Employed a two-step strategy: genome-wide SNP screening using Monte Carlo pedigree disequilibrium test (MCPDT), followed by eLBL on identified haplotype blocks.
• Incorporated familial correlation adjustments for both case-control and trio data.

Main Results:

• Identified several significantly associated haplotypes using the eLBL method.
• Most associated haplotypes were located within protein-coding genes relevant to metabolic syndrome.
• Type I error study and quantitative trait analysis supported the findings.

Conclusions:

• eLBL effectively increases statistical power for detecting rare haplotype associations with complex diseases by jointly analyzing diverse genetic data.
• The identified haplotypes in protein-coding genes suggest potential genetic underpinnings for metabolic syndrome.
• The eLBL methodology offers a robust framework for genetic association studies incorporating familial data.