APRICOT

Admixed Population poweR Inference Computed for phenOtypic Traits

This package provides a mechanistic simulation framework to infer statistical power for discovery of causal variants in an admixed population and the correpsonding ancestral origin populations. It sets in a 2-way admixture scenario with assumption of independent markers, and allows for flexibility in adjusting environmental effect, trait divergence, polygenicity, heritability etc. APRICOT also provides a function to estimate power for phenotype-ancestry association for dichotomous traits based on incidence rates and ancestry distributions. See Lin et al. Front Genet, 2021 for more details on the model.

Setup

In R, you can install APRICOT with the following commands:

devtools::install_github("menglin44/APRICOT")
library(APRICOT)

Functions and Usage

• Genotype-mediated simulation framework to test for GWAS power in an admixed group
  • See ?SimGenoPower
• Power for phenotype-ancestry associations
  • See ?powphenanc
  • This function requires an example ancestry matrix as part of input. If no emprical estimates are available, one can generate an approximate ancestry under a beta distribtuion. A helper function AncBetaOut is provided.

Reference

Lin et al. Front Genet, 2021