Methods for detecting associations between phenotype and aggregations of rare variants

Area of Science:

• Genetics
• Statistical Genetics
• Genomic Association Studies

Background:

• Genome-wide association studies (GWAS) identify genetic variants for numerous traits.
• However, identified variants explain only a small fraction of phenotypic variation, leaving a 'missing heritability' component.
• Rare genetic variants are a leading hypothesis for this missing heritability.

Purpose of the Study:

• To investigate the feasibility of detecting associations from aggregated rare variants across broader genomic regions.
• To extend existing methods, typically focused on short candidate gene regions, to agnostic analysis of entire chromosomes or exomes.
• To identify subsets of variant sites associated with traits using computational approaches.

Main Methods:

• Developed a method to search for significant subsets of variant sites within large genomic regions.
• Employed Markov chain Monte Carlo (MCMC) and genetic algorithms for subset searching.
• Applied the method to analyze genetic data, utilizing known answers from the Genetic Analysis Workshop 17 (GAW17).

Main Results:

• The study demonstrates the potential for identifying associations within broader genomic regions by searching for specific variant subsets.
• The computational methods (MCMC, genetic algorithms) are shown to be feasible for this agnostic, large-scale analysis.
• The analysis successfully utilized GAW17 data to test the proposed methodology.

Conclusions:

• Extending rare variant association analysis to broader genomic regions is feasible.
• Agnostic searching for variant subsets using MCMC or genetic algorithms can help uncover the genetic basis of missing heritability.
• This approach offers a promising avenue for identifying complex genetic contributions to phenotypic variation.