How is permutation done?

[Fernando Andrade]

Dear Karl,

I have been trying to use qtl2 in 2 populatios we have, an AIL and the DO. Both populations heave non-independent individuals, therefore I must use kinship to correctly map the QTL and for this I am using the scan1perm() to get the empirical significance threshold. I was wondering how is this permutation done since these individuals are not independent of each other. Do you use any kind of Gene dropping? (I think not, since gene dropping needs the full pedigree, right?) I was trying to read your qtl2 functions but I could not understand how the permutation (using kinship) is done.

Thank you very much for your attention

Fernando

[Karl Broman]

See the "Details" section in the help file for scan1perm():

"If kinship is not provided, so that analysis proceeds by Haley-Knott regression, we permute the rows of the phenotype data; the same permutations are also applied to the rows of the covariates (addcovar, Xcovar, and intcovar) are permuted.

"If kinship is provided, we instead permute the rows of the genotype data and fit an LMM with the same residual heritability (estimated under the null hypothesis of no QTL)."

karl