Galaxy | Tool Preview

BEAM (version 1.0.0)

This tool can take a long time to run, depending on the number of SNPs, the sample size, and the number of MCMC steps specified. If you have hundreds of thousands of SNPs, it may take over a day. The main tasks that slow down this tool are searching for interactions and dynamically partitioning the SNPs into blocks. Optimization is certainly possible, but hasn't been done yet. If your only interest is to detect SNPs with primary effects (i.e., single-SNP associations), please use the GPASS tool instead.


Dataset formats

The input dataset must be in lped format. The output datasets are both tabular. (Dataset missing?)


What it does

BEAM (Bayesian Epistasis Association Mapping) uses a Markov Chain Monte Carlo (MCMC) method to infer SNP block structures and detect both single-marker and interaction effects from case-control SNP data. This tool also partitions SNPs into blocks based on linkage disequilibrium (LD). The method utilized is Bayesian, so the outputs are posterior probabilities of association, along with block partitions. An advantage of this method is that it provides uncertainty measures for the associations and block partitions, and it scales well from small to large sample sizes. It is powerful in detecting gene-gene interactions, although slow for large datasets.


Example


References

Zhang Y, Liu JS. (2007) Bayesian inference of epistatic interactions in case-control studies. Nat Genet. 39(9):1167-73. Epub 2007 Aug 26.

Zhang Y, Zhang J, Liu JS. (2010) Block-based bayesian epistasis association mapping with application to WTCCC type 1 diabetes data. Submitted.