simGenoFuncDiffPriors: Simulate Genotype Data from a Mixture of 3 Bayesian...

In ubcxzhang/GWASbyCluster: Identifying Significant SNPs in Genome Wide Association Studies (GWAS) via Clustering

Description

Simulate Genotype Data from a Mixture of 3 Bayesian Hierarchical Models. The minor allele frequency (MAF) of cases has different priors than that of controls.

Usage

simGenoFuncDiffPriors(
    nCases = 100, 
    nControls = 100, 
    nSNPs = 1000, 
    alpha.p.ca = 2, 
    beta.p.ca = 3, 
    alpha.p.co = 2, 
    beta.p.co = 8, 
    pi.p = 0.1, 
    alpha0 = 2, 
    beta0 = 5, 
    pi0 = 0.8, 
    alpha.n.ca = 2, 
    beta.n.ca = 8, 
    alpha.n.co = 2, 
    beta.n.co = 3, 
    pi.n = 0.1, 
    low = 0.02, 
    upp = 0.5, 
    verbose = FALSE)

Arguments

nCases	integer. Number of cases.
nControls	integer. Number of controls.
nSNPs	integer. Number of SNPs.
alpha.p.ca	numeric. The first shape parameter of Beta prior in cluster + for cases.
beta.p.ca	numeric. The second shape parameter of Beta prior in cluster + for cases.
alpha.p.co	numeric. The first shape parameter of Beta prior in cluster + for controls.
beta.p.co	numeric. The second shape parameter of Beta prior in cluster + for controls.
pi.p	numeric. Mixture proportion for cluster +.
alpha0	numeric. The first shape parameter of Beta prior in cluster 0.
beta0	numeric. The second shape parameter of Beta prior in cluster 0.
pi0	numeric. Mixture proportion for cluster 0.
alpha.n.ca	numeric. The first shape parameter of Beta prior in cluster - for cases.
beta.n.ca	numeric. The second shape parameter of Beta prior in cluster - for cases.
alpha.n.co	numeric. The first shape parameter of Beta prior in cluster - for controls.
beta.n.co	numeric. The second shape parameter of Beta prior in cluster - for controls.
pi.n	numeric. Mixture proportion for cluster -.
low	numeric. A small positive value. If a MAF generated from half-flat shape bivariate prior is smaller than low, we will delete the SNP to be generated.
upp	numeric. A positive value. If a MAF generated from half-flat shape bivariate prior is greater than upp, we will delete the SNP to be generated.
verbose	logical. Indicating if intermediate results or final results should be output to output screen.

Details

In this simulation, we generate additive-coded genotypes for 3 clusters of SNPs based on a mixture of 3 Bayesian hierarchical models.

In cluster +, the minor allele frequency (MAF) θ_{x+} of cases is greater than the MAF θ_{y+} of controls.

In cluster 0, the MAF θ_{0} of cases is equal to the MAF of controls.

In cluster -, the MAF θ_{x-} of cases is smaller than the MAF θ_{y-} of controls.

The proportions of the 3 clusters of SNPs are π_{+}, π_{0}, and π_{-}, respectively.

We assume a “half-flat shape” bivariate prior for the MAF in cluster +

2h_{+}≤ft(θ_{x+}\right)h_{+}≤ft(θ_{y+}\right) I≤ft(θ_{x+}>θ_{y+}\right),

where I(a) is hte indicator function taking value 1 if the event a is true, and value 0 otherwise. The function h_{+} is the probability density function of the beta distribution Beta≤ft(α_{+}, β_{+}\right).

We assume θ_{0} has the beta prior Beta(α_0, β_0).

We also assume a “half-flat shape” bivariate prior for the MAF in cluster -

2h_{-}≤ft(θ_{x-}\right)h_{-}≤ft(θ_{y-}\right) I≤ft(θ_{x-}>θ_{y-}\right).

The function h_{-} is the probability density function of the beta distribution Beta≤ft(α_{-}, β_{-}\right).

Given a SNP, we assume Hardy-Weinberg equilibrium holds for its genotypes. That is, given MAF θ, the probabilities of genotypes are

Pr(geno=2) = θ^2

Pr(geno=1) = 2θ≤ft(1-θ\right)

Pr(geno=0) = ≤ft(1-θ\right)^2

We also assume the genotypes 0 (wild-type), 1 (heterozygote), and 2 (mutation) follows a multinomial distribution Multinomial≤ft\{1, ≤ft[ θ^2, 2θ≤ft(1-θ\right), ≤ft(1-θ\right)^2 \right]\right\}

Note that when we generate MAFs from the half-flat shape bivariate priors, we might get very small MAFs or get MAFs >0.5. In these cased, we then delete this SNP.

So the final number of SNPs generated might be less than the initially-set number of SNPs.

Value

An ExpressionSet object stores genotype data.

Author(s)

Yan Xu <[e-mail]>, Li Xing <[e-mail]>, Jessica Su <[e-mail]>, Xuekui Zhang <[e-mail]>, Weiliang Qiu <[e-mail]>

References

Yan X, Xing L, Su J, Zhang X, Qiu W. Model-based clustering for identifying disease-associated SNPs in case-control genome-wide association studies. Scientific Reports 9, Article number: 13686 (2019) https://www.nature.com/articles/s41598-019-50229-6.

Examples

set.seed(2)

esSimDiffPriors = simGenoFuncDiffPriors(
  nCases = 100,
  nControls = 100,
  nSNPs = 500,
  alpha.p.ca = 2, beta.p.ca = 3,
  alpha.p.co = 2, beta.p.co = 8, pi.p = 0.1,
  alpha0 = 2, beta0 = 5, pi0 = 0.8,
  alpha.n.ca = 2, beta.n.ca = 8,
  alpha.n.co = 2, beta.n.co = 3, pi.n = 0.1,
  low = 0.02, upp = 0.5, verbose = FALSE
)

print(esSimDiffPriors)

pDat = pData(esSimDiffPriors)
print(pDat[1:2,])
print(table(pDat$memSubjs))

fDat = fData(esSimDiffPriors)
print(fDat[1:2,])
print(table(fDat$memGenes))
print(table(fDat$memGenes2))

ubcxzhang/GWASbyCluster documentation built on Nov. 5, 2019, 11:03 a.m.