This package implements Monte Carlo methods (including efficient importance sampling approaches) for assessing the false positive and false negative rates expected when using a particular set of genetic markers for pairwise relationship inference. It also provides functions implemented in C++ for computing likelihood ratios for different relationships between all pairs of individuals in a data set.
CKMRsim can be installed with the ‘remotes’ packages` It requires that development tools be installed on your system to allow compilation of C++ via RCpp.
remotes::install_github("eriqande/CKMRsim", build_vignettes = TRUE)Once that is done (and note that building the vignettes can take a while), you can read the vignettes with:
vignette("CKMRsim-example-1")
vignette("CKMRsim-example-2-microsatellites")
vignette("CKMRsim-simulating-linked-markers")
vignette("CKMRsim-writing-geno-error-funcs")
vignette("CKMRsim-pairwise-relationships")An easier way to get to read these vignetttes is to find them in the “articles” tab at CKMRsim’s pkgdown page at: https://eriqande.github.io/CKMRsim/
Simulations with physically linked markers are now performed internally using CKMRsim’s Rcpp gene-dropper by default, so Mendel is no longer required for routine linked-marker simulations. The legacy Mendel backend remains available with useMendel = TRUE and install_mendel(), mainly for reproducing older analyses or backend comparisons, but it is no longer the recommended path.
A tutorial on how to use CKMRsim, prepared for a workshop at TWS 2022, can be found at https://eriqande.github.io/tws-ckmr-2022/kin-finding-lab.html