This is an R package that implements a modification of the Random Forests algorithm for analysis of case-control genetic data that corrects the bias of variable importance measures for X chromosome single nucleotide polymorphisms (SNPs). The new method is based on the original Random Forests algorithm developed by Leo Breiman and Adele Cutler, implemented in the R package, randomForest v4.6-7.
The team's modified version prevents bias in variable importance of X chromosome SNPs compared with autosomal SNPs by simulating the process of X chromosome inactivation. The snpRF performs classification based on a forest of trees using random subsets of SNPs, both autosomal and X chromosome, and other variables as inputs.
Available on the Comprehensive R Archive Network (CRAN)