mv_subset: Perform multivariable MR on subset of features
In MRCIEU/TwoSampleMR: Two Sample MR Functions and Interface to MR Base Database

mv_subset

R Documentation

Perform multivariable MR on subset of features

Description

The function proceeds as follows:

Select features (by default this is done using LASSO feature selection).
Subset the mvdat to only retain relevant features and instruments.
Perform MVMR on remaining data.

Usage

mv_subset(
  mvdat,
  features = mv_lasso_feature_selection(mvdat),
  intercept = FALSE,
  instrument_specific = FALSE,
  pval_threshold = 5e-08,
  plots = FALSE
)

Arguments

`mvdat`	Output from `mv_harmonise_data()`.
`features`	Dataframe of features to retain, must have column with name 'exposure' that has list of exposures to retain from mvdat. The default is `mvdat_lasso_feature_selection(mvdat)`.
`intercept`	Should the intercept by estimated (`TRUE`) or force line through the origin (`FALSE`, the default).
`instrument_specific`	Should the estimate for each exposure be obtained by using all instruments from all exposures (`FALSE`, default) or by using only the instruments specific to each exposure (`TRUE`).
`pval_threshold`	P-value threshold to include instruments. The default is `5e-8`.
`plots`	Create plots? The default is `FALSE`.