SimGenoPower: Power Inference for GWAS Discovery in Admixed Populations

In menglin44/APRICOT: Admixed Population poweR Inference Computed for phenOtypic Traits

Power Inference for GWAS Discovery in Admixed Populations

Description

Infers statistical power for discovery of variants in a 2-way admixed group, and the corresponding populations of ancestry origin.

Usage

SimGenoPower(
  n_snps = 1000,
  popsizes = c(1000, 1000, 1000),
  mean_ganc = 0.75,
  sd_ganc = 0.1,
  prop_causal = 0.1,
  hg2 = 0.5,
  fst = 0.2,
  fst_constant = T,
  bg_fst = F,
  freqdist = "unif",
  liability = NA,
  aaf = c(0.001, 0.999),
  envanc = 1,
  envvar = 0,
  alpha = 0.05
)

Arguments

n_snps	Total number of SNPs (causal and non-causal). SNPs are assumed independent from each other. Default is 1000.
popsizes	A vector of sample sizes for the admixed group, Pop 1 (ancestral origin from which the major component of ancestry comes), Pop 2 (ancestral origin from which the minor component of ancestry comes). Default is 1000 each.
mean_ganc	Mean global ancestry from Pop 1 in the admixed group. Default is 0.75.
sd_ganc	Standard deviation of the global ancestry distribution from Pop 1 in the admixed group. Default is 0.1.
prop_causal	Proportion of causal variant among total SNPs. Default is 0.1.
hg2	Heritability of the trait. Default is 0.5.
fst	The baseline fixation index FST among causal variants between Pop 1 and Pop2. Default is 0.2.
fst_constant	True or False: Is FST among causal variants a constant value equaling the baseline (True) or have an increment with the rarity of ancestral minor allele frequency (False)? Default is True.
bg_fst	Genomic background Fst value between 0 and 1 for non-causal variants. Default 'FALSE' to be set as the same as fst. If set to a value here, Fst of non-causal variants will be under this constant value.
freqdist	Frequency distribution of ancestral variants (prior to the divergence of Pop 1 and 2). "unif" for a uniform distribution modeling (default), "inverse" for an inverse distribution modeling.
liability	A vector of liability threshold for dichotomous traits in the admixed group, Pop 1, Pop2, respectively. If set as NA (default), the trait is assumed to be quantitative.
aaf	Limits (min, max) on the frequency distribution of ancestral variants. Default is c(0.001, 0.999).
envanc	Options to model environment by ancestry interactions: 0 means no interaction. 1 means interation is modeled as the sum of Gaussian sampling. 2 means the interaction is modeled as linear dependence on the ancestry proportion. Default is 1.
envvar	percentage of the environment by ancestry interaction over the total phenotypic variance. Only effective when envanc is 2.
alpha	Type I error rate. Default is 0.05.

Value

A list of powers and the re-estimated heritability from simulation.

admixed_power	Power for discovery in the admixed group
pop1_power	Power for discovery in Pop 1
pop2_power	Power for discovery in Pop 2
h2	Estimated heritability from simulated phenotypes and genotypes

menglin44/APRICOT documentation built on July 1, 2023, 4:01 p.m.