HDMT (version 1.0.5)

adjust_quantile: A function to compute the quantiles of the estimated mixture null distribution for pmax using either the approximation or exact method

Description

A function to generate the quantiles of the estimated three-component mixture null distribution for pmax (the maximum of the two p-values for testing mediation) using either approximation or exact method

Usage

adjust_quantile(alpha00, alpha01, alpha10, alpha1, alpha2, 
input_pvalues, exact = 0)

Arguments

alpha00

A numeric number represents the proportion of null \(H_{00}\)

alpha01

A numeric number represents the proportion of null \(H_{01}\)

alpha10

A numeric number represents the proportion of null \(H_{10}\)

alpha1

A numeric number represents the proportion of null alpha==0 (association between exposure and mediator)

alpha2

A numeric number represents the proportion of null beta==0 (association between mediator and outcome adjusted for exposure)

input_pvalues

A matrix contains two columns of p-values for candidate mediators. Column 1 is the p-value of testing if a exposure is associated with the mediator (alpha!=0). Column 2 is the p-value of testing if a mediator is associated with the outcome adjusted for exposure(beta!=0)

exact

Use the option to choose from two methods. exact=0: the approximation method without estimating the CDFs when deriving the mixture null distribution; exact=1: the exact method to estimate the CDFs nonparametrically when deriving the mixture null distribution

Value

A vector contains the expected quantiles of p-values based on the estimated mixture null distribution. See Dai et al (2020) for details of how to compute quantiles using the approximation method (exact=0) or the exact method (exact=1).

Details

The is a function to computed the expected quantiles for the observed p-max values based on the estimated mixture null distribution. The methodology detail can be found in Dai et al (2020).

References

James Y. Dai, Janet L. Stanford, Michael LeBlanc. A multiple-testing procedure for high-dimensional mediation hypotheses. Journal of the American Statistical Association, 2020, DOI: 10.1080/01621459.2020.1765785.

Examples

# NOT RUN {
data(snp_input)
input_pvalues <- snp_input

#To save time for illustration, we use 10 percent of rows

input_pvalues <- input_pvalues[sample(1:nrow(input_pvalues),size=ceiling(nrow(input_pvalues)/10)),]

nullprop <- null_estimation(input_pvalues)

pnull <- adjust_quantile(nullprop$alpha00,nullprop$alpha01,nullprop$alpha10,nullprop$alpha1,
                         nullprop$alpha2,input_pvalues,exact=0)
# }