falseGrandma: estimate per-comparison error rates

In delomast/gRandma: Inference of Grandparentage With Genetic Markers

estimate per-comparison error rates

Description

estimate per-comparison error rates

Usage

falseGrandma(
  gmaData,
  relationship = c("ssGP", "sP"),
  llrToTest,
  N = 10000,
  seed = sample.int(.Machine$integer.max, 1),
  itersPerMI = NULL,
  errorType = c("falseNegative", "pairwise", "Unrel", "Aunt", "HalfAunt", "ParCous",
    "True_GAunt", "True_Unrel", "True_HGAunt", "True_GpCous", "GAunt_Unrel",
    "HGAunt_Unrel", "GpCous_Unrel", "GAunt", "GAunt_HGAunt", "Gaunt_GpCous", "HGAunt",
    "HGAunt_GpCous", "GpCous"),
  MIexcludeProb = 1e-04,
  maxMissingGenos = NULL,
  method = c("strat", "IS", "test")
)

Arguments

gmaData	the gmaData object containing your baseline populations and potential descendents. This input is created by createGmaInput
relationship	the relationship you want to test for: "ssGP" - single sided grandparentage (a pair of either two maternal grandparents OR two paternal grandparents); "sP" - single parent inference
llrToTest	a vector of llr thresholds to estimate error rates for
N	the number of Monte Carlo samples to take to estimate the error rates (ignored for stratified methods)
seed	a positive integer to use as a seed for random number generation.
errorType	the type of error estimate to make. Options for relationship = "ssGP": "falseNegative", "Unrel", "True_GAunt", "True_Unrel", "True_HGAunt", "True_GpCous", "GAunt_Unrel", "HGAunt_Unrel", "GpCous_Unrel", "GAunt", "GAunt_HGAunt", "Gaunt_GpCous", "HGAunt", "HGAunt_GpCous", "GpCous". Options for relationship "sP": "falseNegative", "Unrel", "Aunt", "HalfAunt", "ParCous". The "pairwise" option is experimental and should not be used.
MIexcludeProb	the maximum probability of exclusion for a true relationship due to Mendelian incompatibilities. If 0, then no filtering is performed based on Mendelian incompatibilities.
maxMissingGenos	the maximum number of missing genotypes a sample can have before you would choose to omit it from analysis. Default is 10% of the total number of loci, rounded up
method	strat for stratified, IS for importance sampling - currently only available for ssGP. Do not use method = "test", this is currently for internal testing and will be removed.
itersperMI	the number of iterations per Mendelian incompatibility, in order of 0, 1, ... (ignored for non-stratified methods)

delomast/gRandma documentation built on March 8, 2024, 2:26 a.m.