library(stringr) df1 <- read.table(text = "ABC|DEF|GHI, ABCD|EFG|HIJK, ABCDE|FGHI|JKL, DEF|GHIJ|KLM, GHI|JKLM|NO|PQRS, BCDE|FGHI|JKL") #extract 1st word before | word(df1$V1,1,sep = "|")