powerMRMA: powerMRMA

In SharonLutz/powerMRMA: This is a power package to examine the mediated path from gene to outcome through an intermediate phenotype

Description

Power function to examine the mediated path from gene to outcome through an intermediate phenotype

Usage

powerMRMA(plot.name = "powerMRMAplot", MethodNames = c("MR.Classical", "MR.Egger", "MR.IVW", "MR.Median", "MA.Imai", "MA.4Way"), 
n = 1000, n.sim = 500, MeasurementError = T, alpha.level = 0.05, legend.include = T, color = T, SEED = 1, nSNP = 4, 
MAF = c(0.2, 0.2, 0.2, 0.2), varM = 1, varY = 1, gamma0 = 0, gammaX = c(0.15, 0.15, 0.15, 0.15), gammaU = 0, beta0 = 0, 
betaM = c(0,0.20, 0.25), betaX = c(0, 0, 0, 0), betaI = c(0, 0, 0, 0), betaU = 0, muU = 0, varU = 1, muME = 0, varME = 1)

Arguments

plot.name	is the name of the plot
MethodNames	is the vector of methods to run
n	is the sample size, (default = 1000)
n.sim	is the number of simulations to run (default = 500)
MeasurementError	is generated on the mediator if T, is not generated if F
alpha.level	is the default alpha level
legend.include	is T for adding a legend to the plot, is F for not adding a legend
color	is T for producing a color plot, is F for producing a plot in grayscale
SEED	sets the seed for the function
nSNP	is the number of SNPS to generate
MAF	is the minor allele frequency
varM	is the variance of the mediator M
varY	is the variance of the outcome Y
gamma0	is the intercept for M
gammaX	is a vector of the associations of X with M
gammaU	is the unmeasured confounding on M
beta0	is the intercept for Y
betaM	is the association of M with Y
betaX	is a vector of the direct effects of X on Y
betaI	is a vector of the interaction effects between X and M on Y
betaU	is the unmeasured confounding on Y
muU	is the mean for U
varU	is the variance for U
muME	is the mean for ME
varME	is the varaince for ME

Value

a matrix of the power of each method to detect an effect of the mediator M on the ouctome Y given that the first SNP is associated with the mediator (i.e. the indirect path) and corresponding plot

References

Thwing A, Ghosh D, Hokanson JE, Lutz SM. (2018) Mediated Paths in Genetic Association Studies: A Comparison of Mendelian Randomization and Mediation Analysis Approaches. (Submitted).

SharonLutz/powerMRMA documentation built on Nov. 2, 2020, 2:32 a.m.