dcgerard/vicar: Various Ideas for Confounder Adjustment in Regression

Implements many methods for accounting for unobserved confounding in linear regression. If control genes are available, then the following methods are implementable: a calibrated version of CATE/RUV4 vruv4(), a Bayesian version of RUV ruvb(), a version of RUV that unifies other versions of RUV ruv3(), and a generalized version of RUV ruvimpute(). If control genes are not available, then MOUTHWASH mouthwash() and BACKWASH backwash() are excellent procedures to use as long as only one covariate is of interest. The methods are described in detail in Gerard and Stephens (2020) <doi:10.1093/biostatistics/kxy029> and Gerard and Stephens (2021) <doi:10.5705/ss.202018.0345>.

install.packages("remotes")
remotes::install_github("dcgerard/vicar")