-
1-reformat-data.R
is the R script that preps datasets insimulated-data
for MB-MDR runs. It writes the reformated train/test datasets inreformatted-data/
. -
2-run-mb-mdr.sh
executesmbmdr-4.4.1-mac-64bits.out
--- the MBMDR-4.4.1 software that can be downloaded here --- and writes results tomb-mdr/results/
. Note that one of the important options that affect computation time is-n
of which default is 1000.
Run this code with ./2-run-mb-mdr.sh
.
There are two result files per dataset: output
that contains the SNP combinations and their corresponding statistics and model
that contains the HLO matrices of all SNP combinations.
-
The
3*.Rmd
files parse the MB-MDR results inmb-mdr/results/
, compute the multilocus risk scores (MRS) and write them tomb-mdr/risks/
.kd
indicates k-way interactions within the SNP combinations. -
mb-mdr/documentation/
contains documentation and papers related to the MB-MDR software.