Package 'medScan'

Large Scale Single Mediator Hypothesis Testing

Description

A collection of methods for large scale single mediator hypothesis testing. The six included methods for testing the mediation effect are Sobel's test, Max P test, joint significance test under the composite null hypothesis, high dimensional mediation testing, divide-aggregate composite null test, and Sobel's test under the composite null hypothesis. Du, J., Zhou, X., Hao, W., Liu, Y., Smith, J. A., & Mukherjee, B. (2022). "Methods for Large-scale Single Mediator Hypothesis Testing: Possible Choices and Comparisons."

Usage

medScan(z.alpha, z.beta, method)

Arguments

z.alpha	the z-test statistic for alpha (exposure effect on the mediator).
z.beta	the z-test statistic for beta (mediator effect on the outcome).
method	the method to use for testing the mediation effect. It should belong to one of the six methods: "Sobel", "MaxP", "JT_comp", "HDMT", "DACT", and "Sobel_comp". (1) Sobel’s test (method = "Sobel"), (2) Max P test (method = "MaxP"), (3) joint significance test under the composite null hypothesis (method = "JT_comp"), (4) high dimensional mediation testing (method = "HDMT"), (5) Divide-Aggregate Composite-null Test (method = "DACT"), and (6) Sobel’s test under the composite null hypothesis (method = "Sobel_comp").

Details

The available methods are: (1) Sobel’s test (method = "Sobel"), (2) Max P test (method = "MaxP"), (3) joint significance test under the composite null hypothesis (method = "JT_comp"), (4) high dimensional mediation testing (method = "HDMT"), (5) Divide-Aggregate Composite-null Test (method = "DACT"), and (6) Sobel’s test under the composite null hypothesis (method = "Sobel_comp").

We incorporated code from the DACT R package formerly on CRAN. Author: Zhonghua Liu Maintainer: Zhonghua Liu, [email protected].

Value

A list that contains

pvalues:	p-values for all mediators from the chosen method.
pi:	the estimated proportions of the three null cases from the HDMT method. pi00 is the proportion of alpha=beta=0; pi01 is the proportion of alpha=0 and beta!=0; and pi10 is the proportion of alpha!=0 and beta=0.

References

Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in structural equation models. Sociological methodology, 13, 290-312.

MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). A comparison of methods to test mediation and other intervening variable effects. Psychological methods, 7(1), 83.

Huang, Y. T. (2019). Genome-wide analyses of sparse mediation effects under composite null hypotheses. The Annals of Applied Statistics, 13(1), 60-84.

Liu, Z., Shen, J., Barfield, R., Schwartz, J., Baccarelli, A. A., & Lin, X. (2022). Large-scale hypothesis testing for causal mediation effects with applications in genome-wide epigenetic studies. Journal of the American Statistical Association, 117(537), 67-81.

Dai, J. Y., Stanford, J. L., & LeBlanc, M. (2022). A multiple-testing procedure for high-dimensional mediation hypotheses. Journal of the American Statistical Association, 117(537), 198-213.

Du, Jiacong, et al. "Methods for large‐scale single mediator hypothesis testing: Possible choices and comparisons." Genetic Epidemiology 47.2 (2023): 167-184.

Examples

# simulate data under the mixture null
n=10000
u = runif(n,0,1)
z.alpha = z.beta = rep(NA,0)
pi00 = 0.98
pi10 = 0.01
pi01 = 0.01
for(i in 1:n){
  if(u[i]<=pi00){
    z.alpha[i] = rnorm(1, 0, 1)
    z.beta[i] = rnorm(1, 0, 1)
  } else if (u[i]<= pi00+pi10){
    z.alpha[i] = rnorm(1, 1, 1)
    z.beta[i] = rnorm(1, 0, 1)
  } else {
    z.alpha[i] = rnorm(1, 0, 1)
    z.beta[i] = rnorm(1, 1, 1)
  }
}

# obtain p-values

# method = "Sobel", "MaxP", "HDMT", "Sobel_comp", "JT_comp", "DACT"
obj = medScan(z.alpha = z.alpha, z.beta = z.beta, method = "Sobel")
qqman::qq(obj$pvalues,  xlim = c(0,4), ylim = c(0,4), main = "Sobel")

---