Subset dataframe to contain only cells that match a pattern in r - r

Question

I am begginer in R and this is a very simple question, but I can't find the answer.
I would like to select cells in a table that match a particular pattern and exclude everything else.
Example data:
data.t <- data.frame(ColA = c("NARG_ECOLI^Q:103", "NARG_ECOLI^NARG", "SPEB_KLEP7^Q:103"), ColB = c(NA, NA, NA), ColC = c("KLEP7^Q:103", "NARG_ECOLI^KLEP7", NA), ColD = c("RPOC_ENTFA^Q:2", NA, NA), ColE = c("Y1546_STAS1^Q:6", NA, NA))
which generates a table like this:
ColA ColB ColC ColD ColE
1 NARG_ECOLI^Q:103 NA KLEP7^Q:103 RPOC_ENTFA^Q:2 NA
2 NARG_ECOLI^NARG NA NARG_ECOLI^KLEP7 <NA> NA
3 SPEB_KLEP7^Q:103 NA <NA> <NA> NA
I would like to select only cells containing ECOLI. Thus, the desired output would look like this one:
ColA ColC
1 NARG_ECOLI^Q:103 NARG_ECOLI^KLEP7
2 NARG_ECOLI^NARG <NA>
One possible solution is to visually inspect and make the selections in my data, but the actual table has dozens of columns and hundreds of rows. Any help would be greatly appreciated. Thank you in advance!

Answer 1

If you want to return ONLY the items in the data frame that have "ECOLI" in them, then here is a tidyverse approach
library(tidyverse)
filter_all(data.t, any_vars(grepl("ECOLI", .))) %>%
.[map_lgl(., ~any(grepl("ECOLI", .x)))] %>%
map_df(~replace(.x, !grepl("ECOLI", .x), NA_character_))
# A tibble: 2 x 2
ColA ColC
<fctr> <fctr>
1 NARG_ECOLI^Q:103 <NA>
2 NARG_ECOLI^NARG NARG_ECOLI^KLEP7

Answer 2

data.t <- data.t[grepl('ECOLI', data.t$ColA), ]
To obtain a staggered list of all instances of ECOLI in each column of data.t:
out <- lapply(data.t, grep, pattern='ECOLI', value=T)
If you want to drop 0 length entries.
nout <- sapply(out, length)
out <- out[nout > 0]
nout <- nout[nout > 0]
To merge that staggered list into a rectangular object like a data frame is unwise, but:
mapply(c, out, mapply(rep, NA, max(nout)-nout))

Answer 3

I tried solving this using base functions.
# Data
data.t <- data.frame(ColA = c("NARG_ECOLI^Q:103", "NARG_ECOLI^NARG",
"SPEB_KLEP7^Q:103"), ColB = c(NA, NA, NA), ColC = c("KLEP7^Q:103",
"NARG_ECOLI^KLEP7", NA), ColD = c("RPOC_ENTFA^Q:2", NA, NA), ColE =
c("Y1546_STAS1^Q:6", NA, NA), stringsAsFactors = FALSE)
# First wrote a function to check cell value. If value contains
"ECOLI" then value # of cell is retained else value is replaced with NA
findECOLI <- function(x){
ifelse(grepl("ECOLI", x, fixed = TRUE), x, NA)
}
d1 <- sapply(data.t, findECOLI)
#> d1
# ColA ColB ColC ColD ColE
#[1,] "NARG_ECOLI^Q:103" NA NA NA NA
#[2,] "NARG_ECOLI^NARG" NA "NARG_ECOLI^KLEP7" NA NA
#[3,] NA NA NA NA NA
# Now, remove the rows containing only NA
d1 <- d1[rowSums(is.na(d1)) != ncol(d1), ]
#> d1
# ColA ColB ColC ColD ColE
#[1,] "NARG_ECOLI^Q:103" NA NA NA NA
#[2,] "NARG_ECOLI^NARG" NA "NARG_ECOLI^KLEP7" NA NA
# Remove the columns containing only NA
d1 <- d1[, colSums(is.na(d1)) != nrow(d1)]
#Result:
#>d1
# ColA ColC
#[1,] "NARG_ECOLI^Q:103" NA
#[2,] "NARG_ECOLI^NARG" "NARG_ECOLI^KLEP7"