Identification of multiple rare variants associated with a disease

Area of Science:

• Genetics
• Statistical genetics
• Bioinformatics

Background:

• Advances in sequencing technology enable rare variant identification for complex diseases.
• Existing statistical methods struggle with large datasets and complex variant-disease relationships.

Purpose of the Study:

• To develop and apply a novel statistical model for identifying rare variants associated with complex diseases.
• To analyze the Genetic Analysis Workshop 17 data, considering both disease liability and quantitative traits.

Main Methods:

• Applied a zero-inflated Poisson regression model to handle low-frequency exonic variants.
• Grouped 697 subjects into European, Asian, and African ancestries using principal components analysis.
• Analyzed collapsed rare variants per gene, assuming enrichment in affected individuals.

Main Results:

• Identified UGT1A1 associated with disease liability and FLT1 with quantitative trait Q1 in the combined population.
• Found FLT1 and KDR associated with Q1, and VNN1 with Q2 among causal loci.
• No significant gene associations were found for quantitative trait Q4.

Conclusions:

• The developed statistical model is feasible and capable of detecting multiple rare variants influencing disease risk.
• Analysis results varied between combined and ethnic-specific population analyses, highlighting the importance of population stratification.