MISA: Bayesian Model Search and Multilevel Inference for SNP Association Studies. The functions in this package focus on intermediate throughput case-control association studies, where the outcome of interest is often a binary disease state and where the genetic markers have been chosen to capture variation in a set of related genes, such as those involved in a specific biochemical pathway. Given this data, we are interested in addressing two questions: ”To what extent does the data support an overall association between the pathway and outcome of interest?” and ”Which markers or genes are most likely to be driving this association?” To address both of these questions,this package performs a Bayesian model search technique that utilizes Evolutionary Monte Carlo and searches over models including main effects of all genetic markers and marker-specific genetic effects in a computationally efficient manner. The package incorporates functions that perform a marginal screen on the genetic markers, summarize the output of the model search algorithm, including image plots of the models with the highest posterior probability, marginal summaries of SNP and gene inclusion probabilities and Bayes Factors, and global summaries of the posterior probability and Bayes Factor giving evidence of an association in the set of SNPs of interest.
Keywords for this software
References in zbMATH (referenced in 3 articles )
Showing results 1 to 3 of 3.
- Wen, Xiaoquan: Bayesian model selection in complex linear systems, as illustrated in genetic association studies (2014)
- Li, Zhen; Gopal, Vikneswaran; Li, Xiaobo; Davis, John M.; Casella, George: Simultaneous SNP identification in association studies with missing data (2012)
- Wilson, Melanie A.; Iversen, Edwin S.; Clyde, Merlise A.; Schmidler, Scott C.; Schildkraut, Joellen M.: Bayesian model search and multilevel inference for SNP association studies (2010)