General imputation framework.

ruvimpute(
  Y,
  X,
  ctl,
  k = NULL,
  impute_func = em_miss,
  impute_args = list(),
  cov_of_interest = ncol(X),
  include_intercept = TRUE,
  do_variance = FALSE
)

Arguments

Y	A matrix of numerics. These are the response variables where each column has its own variance. In a gene expression study, the rows are the individuals and the columns are the genes.
X	A matrix of numerics. The covariates of interest.
ctl	A vector of logicals of length `ncol(Y)`. If position i is `TRUE` then position i is considered a negative control.
k	The rank of the underlying matrix. Used by `hard_impute` if that is the value of `impute_func`. If not provided, will be estimated by `num.sv`.
impute_func	A function that takes as input a matrix names `Y` that has missing values and returns a matrix called `Yhat` of the same dimension of `Y` with the missing values filled in. If `do_variance = TRUE`, then `impute_func` should also return `sig_diag` --- a vector of column-specific variance estimates. The default is a wrapper for `softImpute`. I provide a few functions in this package. `knn_wrapper` performs k-nearest-neighbors imputation. `missforest_wrapper` performs random forest imputation. `softimpute_wrapper` performs nuclear norm minimization imputation.
impute_args	A list of additional parameters to pass to `impute_func`.
cov_of_interest	A vector of positive integers. The column numbers of the covariates in X whose coefficients you are interested in. The rest are considered nuisance parameters and are regressed out by OLS.
include_intercept	A logical. If `TRUE`, then it will check `X` to see if it has an intercept term. If not, then it will add an intercept term. If `FALSE`, then `X` will be unchanged.
do_variance	A logical. Does `impute_func` also return estimates of the column-specific variances?

Value

beta2hat The estimates of the coefficients of the covariates of interest that do not correspond to control genes.

betahat_long The estimates of the coefficients. Those corresponding to control genes are set to 0.

sebetahat If do_variance = TRUE, then these are the "standard errors" of beta2hat (but not really).

tstats If do_variance = TRUE, then these are the "t-statistics" of beta2hat (but not really).

pvalues If do_variance = TRUE, then these are the "p-values" of tstats (but not really).

References

Gerard, David, and Matthew Stephens. 2021. "Unifying and Generalizing Methods for Removing Unwanted Variation Based on Negative Controls." Statistica Sinica, 31(3), 1145-1166. doi: 10.5705/ss.202018.0345

Author

David Gerard