Robust adjusted profile score

Usage

mr_raps(b_exp, b_out, se_exp, se_out, parameters = default_parameters())

Arguments

b_exp

Vector of genetic effects on exposure.

b_out

Vector of genetic effects on outcome.

se_exp

Standard errors of genetic effects on exposure.

se_out

Standard errors of genetic effects on outcome.

parameters

A list of parameters. Specifically, over.dispersion, loss.function, and shrinkage:

over.dispersion is a logical concerning should the model consider overdispersion (systematic pleiotropy);
loss.function allows using either the squared error loss ("l2") or robust loss functions/scores ("huber" or "tukey");
shrinkage is a logical specifying whether empirically partially Bayes should be used.

The default is parameters=list(overdispersion = TRUE, loss.function = "huber", shrinkage = FALSE).

Value

List with the following elements:

b: MR estimate
se: Standard error
pval: p-value
nsnp: Number of SNPs

Details

This function calls the mr.raps package. Please refer to the documentation of that package for more detail.

References

Qingyuan Zhao, Jingshu Wang, Jack Bowden, Dylan S. Small. Statistical inference in two-sample summary-data Mendelian randomization using robust adjusted profile score. Annals of Statistics, 48, 3, 1742–1769, doi:10.1214/19-AOS1866