The goal of wqspt is to implement a permutation test method for theweighted quantile sum (WQS) regression.
Weighted quantile sum regression is a statistical technique toevaluate the effect of complex exposure mixtures on an outcome (Carricoet al. 2015). It is a single-index method which estimates a combinedmixture sum effect as well as weights determining each individualmixture component’s contributions to the sum effect. However, the modelfeatures a statistical power and Type I error (i.e., false positive)rate trade-off, as there is a machine learning step to determine theweights that optimize the linear model fit. If the full data is used toestimate both the mixture component weights and the regressioncoefficients, there is high power but also a high false positive ratesince coefficient p-values are calculated for a weighted mixtureindependent variable calculated using weights that have already beenoptimized to find a large effect.
This package provides an alternative method based on a permutationtest that should reliably allow for both high power and low falsepositive rate when utilizing the WQSr. The permutation test is a methodof obtaining a p-value by simulating the null distribution throughpermutations of the data. The permutation test algorithm is describedmore in detail inDay etal. 2022.
You can install this package fromCRAN with:
install.packages("wqspt")You can also install the development version of wqspt fromGitHub with:
# install.packages("devtools")devtools::install_github("drewdstat/wqspt")Here is abrieftutorial vignette on how to use the wqspt package.