Genetic association analysis for common variants in the Genetic Analysis Workshop 18 data: a Dirichlet regression approach.

ABSTRACT: We propose a genetic association analysis using Dirichlet regression to analyze the Genetic Analysis Workshop 18 data. Clinical variables, arranged in a longitudinal data structure, are employed to fit a multistate transition model in which the transition probabilities are served as a response in the proposed analysis. Furthermore, a gene-based association analysis via penalized regression is implemented using the markers at a single-nucleotide polymorphism level that we previously identified via nonpenalized Dirichlet regression.

SUBMITTER: Espin-Garcia O 

PROVIDER: S-EPMC4143809 | biostudies-literature | 2014

REPOSITORIES: biostudies-literature