Tool:MTMCSKAT package for GWAS with efficient resampling, parallelization

Question

[michael.nagle]

Greetings,

Many of you are likely familiar with the popular SKAT method for GWAS. I'm not affiliated with the developers of the wonderful SKAT or MetaSKAT packages, but wanted to share our laboratory's extension that some researchers may find useful.

Multi-threaded Monte Carlo SKAT (MTMC-SKAT) is an R package built on top of SKAT, providing support for a few useful features:

• Efficient parallelization, across cores on a CPU, as well as across nodes on a high-performance cluster
• Adaptive resampling: p-values computed out to a desired number of significant figures and no further, to help make the most of limited computational resources
• Ability to compute very low empirical p-values, for which we otherwise would not have enough RAM

MTMC-SKAT is available on GitHub: https://github.com/naglemi/mtmcskat