rmdr.fast {Rmdr} | R Documentation |
rmdr.fast(R function based on C programs of the MDR approach ): Detection and Characterization of Gene-Gene Interactions and Gene-Environment Interactions in Genetic and Epidemiological Studies. MDR is a nonparametric and genetic model-free alternative to logistic regression for detecting and characterizing nonlinear interactions among discrete genetic and environmental attributes.
rmdr.fast(data, cvk, nbr, ratio = NULL, randomize = FALSE)
data |
the data to be used. The fisrt column describes the class variable (0 for unaffected and 1 for affected). The other columns describe the attributes with values 0,1,2. |
cvk |
The number of cross-validation > 1 |
nbr |
the number of attribute combinations to be considered |
ratio |
the threshold that is used to pool attribute levels into two groups |
randomize |
randomly order the data |
A matrix giving the best combination of loci, the misclassification error and the prediction error.
Mounir Aout, contributions from C. Wachter
Hahn LW, Ritchie MD, Moore JH.: Multifactor dimensionality reduction software for detecting gene-gene and gene-environment interactions.Bioinformatics. 2003 Feb 12;19(3):376-82
data(mdrdata) cvk<-10 nbr=3 res<-rmdr.fast(mdrdata,10,3, randomize=TRUE)