I never remember regex patterns, so here is a common one to separate a column by the last underscore in the name. It’s handy to use when the column names contain information to be stored in a separate variable, for example time (_PRE, _POST, _FU).
separate(col, into = c("col1", "col2"), sep = "_(?=[^_]+$)")