Two-phase and family-based designs for next-generation sequencing studies.

Area of Science:

• Genetics
• Epidemiology
• Bioinformatics

Background:

• Next-generation sequencing (NGS) costs are decreasing, approaching Genome-Wide Association Study (GWAS) panel prices.
• Challenges remain for large epidemiologic studies due to sequencing costs and distinguishing causal from non-causal rare variants.

Purpose of the Study:

• To review and compare two primary designs for sequencing studies: two-phase designs and family-based designs.
• To evaluate strategies for variant prioritization and association testing in the context of rare variants and large datasets.

Main Methods:

• Two-phase designs: Subsampling subjects from a larger case-control study for targeted sequencing, followed by association testing.
• Family-based designs: Utilizing co-segregation of variants within families to prioritize causal variants and improve genotype imputation.
• Comparison of relative efficiencies for variant discovery and prioritization between designs using unrelated and related individuals.

Main Results:

• Two-phase designs integrate full sequence data with SNP data for robust association testing.
• Family-based designs, while less efficient for initial discovery, excel at distinguishing causal variants through co-segregation analysis.
• Generalizations of burden or kernel tests show promise for gene-level associations in family-based studies.

Conclusions:

• Both two-phase and family-based designs offer valuable strategies for genetic studies, each with distinct advantages for variant discovery and prioritization.
• Family-based designs are particularly effective for prioritizing causal variants and can be enhanced by family history and identity-by-descent information.
• Advanced statistical methods are crucial for maximizing the power of these designs, especially for rare variants and gene-level associations.