R extract genes from Biostrings format within a list

Question

I have a list of sequences

class(myseq)
[1] "list"

Inside of each element of the list are stores multiples elements withe the format of the Biostrings package. In this example the list myseq contain 6 samples

names(myseq)
 [1] "Sample-01" "Sample-02" "Sample-03" "Sample-04" "Sample-05" "Sample-06"

and each sample have a format of the Biostrings package

class(myseq[["Sample-01"]])
[1] "AAStringSet"
attr(,"package")
[1] "Biostrings"

myseqs[["Sample-01"]]
AAStringSet object of length 143:
      width seq                                                                                                                      names               
  [1]   453 MISIIKRFLGKRQPRQSAEHHYEFLPAHLALAQKPPSPFARLTAITLSIGVLAVLLWAY...VFPAQVQLNKNNIVIDGQTVELTPGMSVVAEIKTDKRRVIDYLLSPIREYQAEALRER IMEHDJCA_00190
  [2]   701 MSNNEGLTLICIHFYLSIISGNREQFKENANLTKTNNYKEELKKIQKANKVRITTKHSQ...VITIAHRLSTVRDCNRIIVLHQGAIVEQGSHQQLLTHGKQYKQLWQLQQELKQEETTA IMEHDJCA_00191
  ...   ... ...
[142]   275 MTAITISDQEYRDFSRFLESQCGIVLGDSKQYLVRSRLSPLVTKFKLASLSDLLRDVVT...RNVLIYFSPDMKSKVLNQMANSLNPGGYLLLGASESLTGLTDRFEMVRCNPGIIYKLK IMEHDJCA_03929
[143]   172 MPLLDSFTVDHTRMHAPAVRVAKTMQTPKGDTITVFDLRFTAPNKDILSEKGIHTLEHL...ESQNKIPELNEYQCGTAAMHSLDEAKQIAQNILAVGISVNRNDELALPEAMLKELKVD IMEHDJCA_04048

so I want to extract an specif gene of each sample using a data.frame

head(df)
           qseqid   samples gene
1  IMEHDJCA_02683 Sample-01 pilB
2  DIBHEKPI_01114 Sample-02 pilB
3  LLMDBGDK_00899 Sample-03 pilB
4  EBMGLGMO_01529 Sample-04 pilB
5  ILCJGNBA_00973 Sample-05 pilB
6  JAGNDBFC_01143 Sample-06 pilB

Extract the genes of each Sample using the qseqid column with a for loop

genes <- c()
for(i in 1:nrow(df)){
    sample <- df[i,2]
    qseq_id <- df[i,1]

    seq <- myseqs[[sample]]
    genes[[i]] <- Biostrings::AAStringSet(seq[qseq_id])
}

my problem is that the gene variable is a list

[[1]]
AAStringSet object of length 1:
    width seq                                                                                                                        names               
[1]   562 MQSNLATILRQANQLSLTQEQACRETIQASGVTAPEALLQLGFFQPDELTEKLSAIFGLP...YEVMPFDEQLAEAIVKGASVQSLEMLARQKGMMTLKDSGLEKLKQGITSLEELQRVLYL IMEHDJCA_02683

[[2]]
AAStringSet object of length 1:
    width seq                                                                                                                        names               
[1]   562 MQSNLATILRQANQLSLTQEQACRETIQASGVTAPEALLQLGFFQPDELTEKLSAIFGLP...YEVMPFDEQLAEAIVKGASVQSLEMLARQKGMMTLKDSGLEKLKQGITSLEELQRVLYL DIBHEKPI_01114

[[3]]
AAStringSet object of length 1:
    width seq                                                                                                                        names               
[1]   561 MTNLATILRQANQLSLTQEQACRETIQASGVTAPEALLQLGFFQPDELTEQLSAIFGLPC...YEVMPFDEQLAEAIVKGASVQSLEMLAQQKGMMTLKDSGLEKLKQGITSLEELQRVLYL LLMDBGDK_00899

[[4]]
AAStringSet object of length 1:
    width seq                                                                                                                        names               
[1]   562 MQSNLATILRQANQLSLTQEQACRETIQASGVTAPEALLQLGFFQPDELTEKLSAIFGLP...YEVMPFNEQLAEAIVKGASVQSLEMLARQKGMMTLKDSGLEKLKQGITSLEELQRVLYL EBMGLGMO_01529

[[5]]
AAStringSet object of length 1:
    width seq                                                                                                                        names               
[1]   562 MQSNLATILRQANQLSLTQEQACRETIQASGVTAPEALLQLGFFQPDELTEKLSAIFGLP...YEVMPFNEQLAEAIVKGASVQSLEMLARQKGMMTLKDSGLEKLKQGITSLEELQRVLYL ILCJGNBA_00973

[[6]]
AAStringSet object of length 1:
    width seq                                                                                                                        names               
[1]   562 MQSNLATILRQANQLSLTQEQACRETIQASGVTAPEALLQLGFFQPDELTEKLSAIFGLP...YEVMPFNEQLAEAIVKGASVQSLEMLARQKGMMTLKDSGLEKLKQGITSLEELQRVLYL JAGNDBFC_01143

And I want just a single Biostrings object that contain the genes of the 6 samples, something like:

genes
AAStringSet object of length 6:
    width seq                                                                                                                                                                                                                                                                                       names               
[1]   562 MQSNLATILRQANQLSLTQEQACRETIQASGVTAPEALLQLGFFQPDELTEKLSAIFGLP...YEVMPFDEQLAEAIVKGASVQSLEMLARQKGMMTLKDSGLEKLKQGITSLEELQRVLYL IMEHDJCA_02683
[2]   562 MQSNLATILRQANQLSLTQEQACRETIQASGVTAPEALLQLGFFQPDELTEKLSAIFGLP...YEVMPFDEQLAEAIVKGASVQSLEMLARQKGMMTLKDSGLEKLKQGITSLEELQRVLYL DIBHEKPI_01114
[3]   561 MTNLATILRQANQLSLTQEQACRETIQASGVTAPEALLQLGFFQPDELTEQLSAIFGLPC...YEVMPFDEQLAEAIVKGASVQSLEMLAQQKGMMTLKDSGLEKLKQGITSLEELQRVLYL LLMDBGDK_00899
[4]   562 MQSNLATILRQANQLSLTQEQACRETIQASGVTAPEALLQLGFFQPDELTEKLSAIFGLP...YEVMPFNEQLAEAIVKGASVQSLEMLARQKGMMTLKDSGLEKLKQGITSLEELQRVLYL EBMGLGMO_01529
[5]   562 MQSNLATILRQANQLSLTQEQACRETIQASGVTAPEALLQLGFFQPDELTEKLSAIFGLP...YEVMPFNEQLAEAIVKGASVQSLEMLARQKGMMTLKDSGLEKLKQGITSLEELQRVLYL ILCJGNBA_00973
[6]   562 MQSNLATILRQANQLSLTQEQACRETIQASGVTAPEALLQLGFFQPDELTEKLSAIFGLP...YEVMPFNEQLAEAIVKGASVQSLEMLARQKGMMTLKDSGLEKLKQGITSLEELQRVLYL JAGNDBFC_01143

I have tried to store the gene variable without [[i]]:

genes <- Biostrings::AAStringSet(seq[qseq_id])

but it only keep the last sequence.

If I do it manually, will be like:

S01 <- Biostrings::AAStringSet(myseqs[["Sample-01"]]["IMEHDJCA_02683"])
S02 <- Biostrings::AAStringSet(myseqs[["Sample-02"]]["DIBHEKPI_01114"])

and so on for Sample-03 to Sample-n.... and then generate the Biostrings object

genes <- c(S01, S02, S03, S04, S05, S06) 

Does any body know how to do it???

Thanks So Much !!!

Answer 1

You can unlist the AAStringSet object, eg

library(Biostrings)

a <- AAString("MISIIKRFLGKRQPRQSAEHHYEFLPAHLALAQKPPSPFARLTAITLSIGVLAVLLWAYVFPAQVQLNKNNIVIDGQTVELTPGMSVVAEIKTDKRRVIDYLLSPIREYQAEALRER")
b <- AAString("MTAITISDQEYRDFSRFLESQCGIVLGDSKQYLVRSRLSPLVTKFKLASLSDLLRDVVTRNVLIYFSPDMKSKVLNQMANSLNPGGYLLLGASESLTGLTDRFEMVRCNPGIIYKLK")
myseqs <- list("Sample-01" = AAStringSet(c("IMEHDJCA_02683" = a)),
               "Sample-02" = AAStringSet(c("DIBHEKPI_01114" = b)))
class(myseqs)
#> [1] "list"
names(myseqs)
#> [1] "Sample-01" "Sample-02"
myseqs[["Sample-01"]]
#> AAStringSet object of length 1:
#>     width seq                                               names               
#> [1]   117 MISIIKRFLGKRQPRQSAEHHYE...KRRVIDYLLSPIREYQAEALRER IMEHDJCA_02683

class(myseqs[["Sample-01"]])
#> [1] "AAStringSet"
#> attr(,"package")
#> [1] "Biostrings"

df <- read.table(text = "       qseqid   samples gene
1 IMEHDJCA_02683 Sample-01 pilB
2 DIBHEKPI_01114 Sample-02 pilB", header = TRUE)
df
#>           qseqid   samples gene
#> 1 IMEHDJCA_02683 Sample-01 pilB
#> 2 DIBHEKPI_01114 Sample-02 pilB

genes <- c()
for(i in 1:nrow(df)){
  sample <- df[i,2]
  qseq_id <- df[i,1]
  
  seq <- myseqs[[sample]]
  genes[[i]] <- unlist(AAStringSet(seq[qseq_id]))
}

genes_object <- AAStringSet(genes)

genes_object
#> AAStringSet object of length 2:
#>     width seq
#> [1]   117 MISIIKRFLGKRQPRQSAEHHYEFLPAHLALAQK...MSVVAEIKTDKRRVIDYLLSPIREYQAEALRER
#> [2]   117 MTAITISDQEYRDFSRFLESQCGIVLGDSKQYLV...GGYLLLGASESLTGLTDRFEMVRCNPGIIYKLK

^{Created on 2022-08-09 by the reprex package (v2.0.1)}

But, unfortunately, a side effect is that you lose the @ranges@NAMES attribute (ie the "names" column disappears).

---

I think you can add the names to the new object, but you would need to check whether it works for your use-case (ie check that the result is correct):

library(Biostrings)

a <- AAString("MISIIKRFLGKRQPRQSAEHHYEFLPAHLALAQKPPSPFARLTAITLSIGVLAVLLWAYVFPAQVQLNKNNIVIDGQTVELTPGMSVVAEIKTDKRRVIDYLLSPIREYQAEALRER")
b <- AAString("MTAITISDQEYRDFSRFLESQCGIVLGDSKQYLVRSRLSPLVTKFKLASLSDLLRDVVTRNVLIYFSPDMKSKVLNQMANSLNPGGYLLLGASESLTGLTDRFEMVRCNPGIIYKLK")
myseqs <- list("Sample-01" = AAStringSet(c("IMEHDJCA_02683" = a)),
               "Sample-02" = AAStringSet(c("DIBHEKPI_01114" = b)))
class(myseqs)
#> [1] "list"
names(myseqs)
#> [1] "Sample-01" "Sample-02"
myseqs[["Sample-01"]]
#> AAStringSet object of length 1:
#>     width seq                                               names               
#> [1]   117 MISIIKRFLGKRQPRQSAEHHYE...KRRVIDYLLSPIREYQAEALRER IMEHDJCA_02683

class(myseqs[["Sample-01"]])
#> [1] "AAStringSet"
#> attr(,"package")
#> [1] "Biostrings"

df <- read.table(text = "       qseqid   samples gene
1 IMEHDJCA_02683 Sample-01 pilB
2 DIBHEKPI_01114 Sample-02 pilB", header = TRUE)
df
#>           qseqid   samples gene
#> 1 IMEHDJCA_02683 Sample-01 pilB
#> 2 DIBHEKPI_01114 Sample-02 pilB

genes <- c()
qseqids <- c()
for(i in 1:nrow(df)){
  sample <- df[i,2]
  qseq_id <- df[i,1]
  
  seq <- myseqs[[sample]]
  genes[[i]] <- unlist(AAStringSet(seq[qseq_id]))
  qseqids <- c(qseqids, qseq_id)
}
names(genes) <- qseqids

genes_object <- AAStringSet(genes)
genes_object
#> AAStringSet object of length 2:
#>     width seq                                               names               
#> [1]   117 MISIIKRFLGKRQPRQSAEHHYE...KRRVIDYLLSPIREYQAEALRER IMEHDJCA_02683
#> [2]   117 MTAITISDQEYRDFSRFLESQCG...SLTGLTDRFEMVRCNPGIIYKLK DIBHEKPI_01114

^{Created on 2022-08-09 by the reprex package (v2.0.1)}