|
Abstract:
|
The state of the art genome-wide association study(GWAS) under the linear mixed model framework, although vastly improved, still suffers from high computational cost. Newer approaches like EMMA, GEMMA and EMMAX among others are better when compared to the traditional GWAS approach, but can still be computationally slow. We propose a new two step mixed model approach, calling it the Randomized Fixed Model(RFM) method, which can be applied under the linear mixed model GWAS framework. Our analyses show that the RFM approach's variance estimate is comparable to the traditional GWAS method while being computationally more efficient.
|
