Statistics – Applications
Scientific paper
2012-03-14
Statistics
Applications
14 pages, 2, figures, 3 tables
Scientific paper
Motivation: Identification of causal SNPs in most genome wide association studies relies on approaches that consider each SNP individually. Increasingly, modern regression methods are employed for SNP selection that consider all markers simultaneously and thus take account of dependencies among SNPs. A promising multivariate strategy for prioritizing biomarkers are CAR scores but so far their application has been restricted to moderately sized data sets. Results: For estimating CAR scores from high-dimensional data we introduce a novel computationally efficient procedure. Subsequently, we conduct a comprehensive comparison study comprising five advanced regression approaches (boosting, lasso, NEG, MCP, and CAR score regression) and a univariate approach (marginal correlation) to find true causal SNPs in data published by the GAW17 consortium. Using this algorithm we demonstrate that SNP rankings based on CAR scores consistently outperform all competing approaches, both uni- and multivariate. Availability: R code to replicate the complete analysis is available from http://strimmerlab.org/software/care/
Duarte Silva Pedro A.
Strimmer Korbinian
Zuber Verena
No associations
LandOfFree
An efficient approach to simultaneous SNP selection: A case study on GAW17 data does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with An efficient approach to simultaneous SNP selection: A case study on GAW17 data, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and An efficient approach to simultaneous SNP selection: A case study on GAW17 data will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-715750