ds.exwas: Performs a non-disclosive EXposome-Wide Association Study

In isglobal-brge/dsExposomeClient: DataSHIELD client site Exposome functions

Performs a non-disclosive EXposome-Wide Association Study

Description

Takes as input an Exposome Set object on the study server and performs a ExWAS association study with the provided model.

Usage

ds.exwas(
  model,
  Set,
  family,
  type = c("pooled", "meta"),
  exposures_family = NULL,
  adjust.by.study = FALSE,
  tef = TRUE,
  datasources = NULL
)

Arguments

model	character Formula, not including exposures, to be tested.
Set	character Name of the Exposome Set object on the server side
family	character Nature of the health outcome (gaussian, binomial or poisson)
type	character Type of analysis to be performed. "pooled" to perform a pooled analysis as if all the data was present on the same dataset, "meta" to perform an ExWAS on every study server to later meta-analyse the results.
exposures_family	character (default NULL) Family to subset the ExposomeSet, only the exposures of the selected family will be used for the ExWAS analysis.
adjust.by.study	bool (default FALSE) Option to adjust the linear models by cohort, this only applies to type = "pooled", there is no need to have a specific variable on the ExposomeSets that has the cohort code. The translation to typical glm syntaxis is objective_variable ~ covariates + cohort (IMPORTANT to note that there is no need to change your model argument of the ds.exwas() function).
tef	bool If TRUE computes the threshold for effective tests.
datasources	a list of DSConnection-class objects obtained after login

Value

For type='pooled': list that contains:
- exwas_results: data.frame with exposure name, coefficient and p-value of the association
- alpha_corrected: numeric effective tests
For type='meta': A list of list with the elements from type='pooled'.

Examples

## Not run: Refer to the package Vignette for examples.

isglobal-brge/dsExposomeClient documentation built on March 5, 2024, 12:26 p.m.