R, dplyr split dataframe by string in columns

Question

I've seen examples here that show how to split a dataframe using column index ranges but how do I split my dataframe with dplyr using strings found in the column? I've purposely created my dataframes so the columns have certain strings in them for future splitting.

Example data:

Site   A_Argas   A_Arachnicea   A_Brus   B_Burkoll   B_Brielle  B_Bact
 1       10        0              0         0           0         0
 2       0         0              0         10          22        123
 3       1         2              3         88          12        546

I want to split this dataframe up based on strings such as "A_" or "B_" and assign them to new dataframes.

For example the output would be:

 dataframeA
 Site   A_Argas   A_Arachnicea   A_Brus 
 1       10        0              0        
 2       0         0              0        
 3       1         2              3       

 dataframeB
 Site   B_Burkoll   B_Brielle  B_Bact
  1       0           0         0
  2      10          22        123
  3      88          12        546

Because this data is not in long format, I can't seem to figure how to change my old code that I used to split the longform dataframes (for a different analysis).

dataframeA <- data %>% filter("GroupID" == "Arachnids") # where "A_" in column headers signify arachnid species
dataframeB <- data %>% filter("GroupID" == "Bacteria") # where all "B_" in the column headers are bacterial species

Answer 1

One base R option might be to grep for the columns starting with A_ and B_, then to subset your original data frame.

a_names <- grep("^A_", names(data), value=TRUE)
dataframeA <- data[ , c("Site", a_names)]
b_names <- grep("^B_", names(data), value=TRUE)
dataframeB <- data[ , c("Site", b_names)]