sim_phenos: Generate phenotypes for three populations YRI, CEU, and CHB.
In KaiqianZhang/prsim: Simulations for PRS Analysis

Generate phenotypes for three populations YRI, CEU, and CHB.

sim_phenos(
  seed = 1234,
  geno_mat,
  chunk_size,
  cores_used,
  af_vec,
  N_YRI,
  N_CEU,
  N_CHB,
  p_YRI,
  p_CEU,
  p_CHB,
  h2_YRI,
  h2_CEU,
  h2_CHB,
  prare_YRI,
  prare_CEU,
  prare_CHB,
  dist,
  ld
)

`geno_mat`	a BEDMartix of dimension N (number of individuals from all three populations) by M (number of SNPs).
`chunk_size`	the size of chunk we perform matrix multiplication
`cores_used`	the number of cores used in parallel computing
`N_YRI`	the number of YRI individuals
`N_CEU`	the number of CEU individuals
`N_CHB`	the number of CHB individuals
`p_YRI`	the proportion of causal SNPs in YRI population
`p_CEU`	the proportion of causal SNPs in CEU population
`p_CHB`	the proportion of causal SNPs in CHB population
`h2_YRI`	heritability of this phenotype in YRI population
`h2_CEU`	heritability of this phenotype in CEU population
`h2_CHB`	heritability of this phenotype in CHB population
`prare_YRI`	proportion of causal rare SNPs in YRI population
`prare_CEU`	proportion of causal rare SNPs in CEU population
`prare_CHB`	proportion of causal rare SNPs in CHB population
`dist`	distribution of effect sizes, default is "gaussian". Options can be "gaussian", "exponential", "laplace". Note that LDpred simulates non-zero effect sizes from laplace distribution.
`ld`	linkage disequilibrium. Options can be "strong" or "weak". We simulate nearby snps together when ld is strong. We simulate snps evenly spread over all variants when ld is weak.
`maf_vec`	the vector of MAF for all SNPs