Remove string in parenthesis and add that as a new column - regex

Question

Possible duplicate Here
I have a data frame of two columns. I want to remove the string in parenthesis and add that as a new column. Data frame is displayed below.
structure(list(ID = 1:12, Gene.Name = structure(c(3L, 11L, 9L,
5L, 1L, 8L, 2L, 4L, 6L, 12L, 10L, 7L), .Label = c(" ATP synt, H+ tran, O subunit (oligomycin sensitivity conferring protein) (ATP5O), mRNA",
" heterogeneous nuclear ribonucleoprotein F (HNRPF), mRNA", " NADH (ubiquinone) 1 alpha subcomplex, 4 (9kD, MLRQ) (NDUFA4), mRNA",
" ribosomal protein L34 (RPL34), transcript variant 1, mRNA",
" ribosomal protein S11 (RPS11), mRNA", "ATP synthase, H+ tran, mitochondrial F0, subunit c (subunit 9) isoform 3 (ATP5G3), mRNA",
"clone MGC:10120 IMAGE:3900723, mRNA, complete cds", "cytidine monophosphate N-acetylneuraminic acid synthetase (CMAS), mRNA",
"farnesyl-diphosphate farnesyltransferase 1 (FDFT1), mRNA", "homeobox protein from AL590526 (LOC84528), mRNA",
"mitochondrial S33 (MRPS33), transcript variant 1, nuclear gene, mRNA",
"ribosomal protein S15a (RPS15A), mRNA"), class = "factor")), .Names = c("ID",
"Gene.Name"), row.names = c(NA, -12L), class = "data.frame")
if the string in parenthesis is not found, then leave that row empty. Here i have two cases
1) Get all the string in parenthesis and add as a new column separated by ,
2) Last string in parenthesis and add as new column
I tried something like df$Symbol <- sapply(df, function(x) sub("\\).*", "", sub(".*\\(", "", x))) but does not give the appropriate output
Case 1 output
ID Gene.Name Symbol
1 NADH (ubiquinone) 1 alpha subcomplex, 4 (9kD, MLRQ) (NDUFA4), mRNA ubiquinone, (9kD, MLRQ),NDUFA4
2 mitochondrial S33 (MRPS33), transcript variant 1, nuclear gene, mRNA MRPS33
3 farnesyl-diphosphate farnesyltransferase 1 (FDFT1), mRNA FDFT1
4 ribosomal protein S11 (RPS11), mRNA RPS11
5 ATP synt, H+ tran, O subunit (oligomycin sensitivity conferring protein) (ATP5O), mRNA oligomycin sensitivity conferring protein,ATP5O
6 cytidine monophosphate N-acetylneuraminic acid synthetase (CMAS), mRNA CMAS
7 heterogeneous nuclear ribonucleoprotein F (HNRPF), mRNA HNRPF
8 ribosomal protein L34 (RPL34), transcript variant 1, mRNA RPL34
9 ATP synthase, H+ tran, mitochondrial F0, subunit c (subunit 9) isoform 3 (ATP5G3), mRNA subunit 9,ATP5G3
10 ribosomal protein S15a (RPS15A), mRNA RPS15A
11 homeobox protein from AL590526 (LOC84528), mRNA LOC84528
12 clone MGC:10120 IMAGE:3900723, mRNA, complete cds NA
Case 2 output
ID Gene.Name Symbol
1 NADH (ubiquinone) 1 alpha subcomplex, 4 (9kD, MLRQ) (NDUFA4), mRNA NDUFA4
2 mitochondrial S33 (MRPS33), transcript variant 1, nuclear gene, mRNA MRPS33
3 farnesyl-diphosphate farnesyltransferase 1 (FDFT1), mRNA FDFT1
4 ribosomal protein S11 (RPS11), mRNA RPS11
5 ATP synt, H+ tran, O subunit (oligomycin sensitivity conferring protein) (ATP5O), mRNA ATP5O
6 cytidine monophosphate N-acetylneuraminic acid synthetase (CMAS), mRNA CMAS
7 heterogeneous nuclear ribonucleoprotein F (HNRPF), mRNA HNRPF
8 ribosomal protein L34 (RPL34), transcript variant 1, mRNA RPL34
9 ATP synthase, H+ tran, mitochondrial F0, subunit c (subunit 9) isoform 3 (ATP5G3), mRNA ATP5G3
10 ribosomal protein S15a (RPS15A), mRNA RPS15A
11 homeobox protein from AL590526 (LOC84528), mRNA LOC84528
12 clone MGC:10120 IMAGE:3900723, mRNA, complete cds <NA>

Answer 1

An option using sub to get the words inside the round brackets at the end of the string.
Symbol <- sub('.*\\(([^\\)]+)\\)[^\\(]+$', '\\1',df1[,2])
df1$Symbol <- Symbol[1:nrow(df1)*NA^(!grepl('\\(',df1[,2]))]
df1$Symbol
#[1] "NDUFA4" "MRPS33" "FDFT1" "RPS11" "ATP5O" "CMAS"
#[7] "HNRPF" "RPL34" "ATP5G3" "RPS15A" "LOC84528" NA
Update
For the first case, ie. extract all characters within the round brackets and paste them together using ,, one option is rm_round from qdapRegex. The output of rm_round is a list. So we use lapply/sapply to loop through the list. Strings that have , inside are separated with grep and we paste the round brackets, and then paste the strings together with collapse=', '. A convenient wrapper function is toString.
library(qdapRegex)
df1$allSymbol <- sapply(rm_round(df1[,2],extract=TRUE), function(x) {
indx <- grep(',', x)
x[indx] <-paste0("(", x[indx], ")")
toString(x)})
is.na(df1$allSymbol) <- df1$allSymbol=='NA'
df1[3:4]
# allSymbol Symbol
#1 ubiquinone, (9kD, MLRQ), NDUFA4 NDUFA4
#2 MRPS33 MRPS33
#3 FDFT1 FDFT1
#4 RPS11 RPS11
#5 oligomycin sensitivity conferring protein, ATP5O ATP5O
#6 CMAS CMAS
#7 HNRPF HNRPF
#8 RPL34 RPL34
#9 subunit 9, ATP5G3 ATP5G3
#10 RPS15A RPS15A
#11 LOC84528 LOC84528
#12 <NA> <NA>

Answer 2

I think I took the easy way out, but if you can get away with it, only match the things in the parentheses that look like a gene symbol, ie, only capital letters and digits
dd <- structure(list(ID = 1:12, Gene.Name = structure(c(3L, 11L, 9L, 5L, 1L, 8L, 2L, 4L, 6L, 12L, 10L, 7L), .Label = c(" ATP synt, H+ tran, O subunit (oligomycin sensitivity conferring protein) (ATP5O), mRNA", " heterogeneous nuclear ribonucleoprotein F (HNRPF), mRNA", " NADH (ubiquinone) 1 alpha subcomplex, 4 (9kD, MLRQ) (NDUFA4), mRNA", " ribosomal protein L34 (RPL34), transcript variant 1, mRNA", " ribosomal protein S11 (RPS11), mRNA", "ATP synthase, H+ tran, mitochondrial F0, subunit c (subunit 9) isoform 3 (ATP5G3), mRNA", "clone MGC:10120 IMAGE:3900723, mRNA, complete cds", "cytidine monophosphate N-acetylneuraminic acid synthetase (CMAS), mRNA", "farnesyl-diphosphate farnesyltransferase 1 (FDFT1), mRNA", "homeobox protein from AL590526 (LOC84528), mRNA", "mitochondrial S33 (MRPS33), transcript variant 1, nuclear gene, mRNA", "ribosomal protein S15a (RPS15A), mRNA"), class = "factor")), .Names = c("ID", "Gene.Name"), row.names = c(NA, -12L), class = "data.frame")
dd$Gene.Name <- as.character(dd$Gene.Name)
## case 1
mm <- gregexpr('(?<=\\()(.*?)(?=\\))', dd$Gene.Name, perl = TRUE)
mm <- regmatches(dd$Gene.Name, mm)
dd <- cbind(dd, case1 = sapply(mm, function(x)
ifelse(length(x), paste(x, collapse = ', '), NA)))
dd[, c(1,3)]
# ID case1
# 1 1 ubiquinone, 9kD, MLRQ, NDUFA4
# 2 2 MRPS33
# 3 3 FDFT1
# 4 4 RPS11
# 5 5 oligomycin sensitivity conferring protein, ATP5O
# 6 6 CMAS
# 7 7 HNRPF
# 8 8 RPL34
# 9 9 subunit 9, ATP5G3
# 10 10 RPS15A
# 11 11 LOC84528
# 12 12 <NA>
## case 2
mm <- gregexpr('(?<=\\()([A-Z0-9]+)(?=\\))', dd$Gene.Name, perl = TRUE)
mm <- regmatches(dd$Gene.Name, mm)
dd <- cbind(dd, case2 = sapply(mm, function(x) ifelse(length(x), x, NA)))
dd[, c(1,4)]
# ID case2
# 1 1 NDUFA4
# 2 2 MRPS33
# 3 3 FDFT1
# 4 4 RPS11
# 5 5 ATP5O
# 6 6 CMAS
# 7 7 HNRPF
# 8 8 RPL34
# 9 9 ATP5G3
# 10 10 RPS15A
# 11 11 LOC84528
# 12 12 <NA>