This is a test version of Biostars. For the public version, visit https://www.biostars.org.
create a gene2accession file from a 2 column file

I need to create a gene2accession file where one gene ID might have one or more Uniprot entries. My file looks like this:

  comp100002_c0 Q9FFI3
  comp100004_c0 B9DHK3
  comp100004_c0 F4J3J5
  comp100005_c0 P54150

and I need it to look like this:

  comp100002_c0 Q9FFI3
  comp100004_c0 B9DHK3|F4J3J5   
  comp100005_c0 P54150

I tried a python script but it doesn't work and after tweaking the code around for a while I am quite stuck. Does anyone have a code that works to get this outcome? Thanks

my attempt in python:

  f1 = open(sys.argv[1], 'rU')
  lines = f1.readlines()
  for i in range(0, len(lines)):
      line = lines[i]
      next_l = f1.next()
      splitline = line.split('\t')
      splitnext = next_l.split('\t')
      if splitline[0] == splitnext[0]:
              print splitline[0] + '\t' + splitline[1] + '|' + splitnext[1]
      else:
          print line
go enrichment cluego uniprot gene2accession
• 2,337 views • link
updated by EagleEye EagleEye • written by Illinu Illinu

1 answer

I guess this post doesn't fits this forum but I hope this solution serves your purpose.

Group by first column using AWK,

awk 'BEGIN{FS="\t"}{ if( !seen[$1]++ ) order[++oidx] = $1; stuff[$1] = stuff[$1] $2 "| " } END { for( i = 1; i <= oidx; i++ ) print order[i]"\t"stuff[order[i]] }' FILE_INPUT_TAB_SEPARATED

OUTPUT

comp100002_c0   Q9FFI3|
comp100004_c0   B9DHK3| F4J3J5|
comp100005_c0   P54150|

If you would like without last occurrence of '|' from the output, use this

awk 'BEGIN{FS="\t"}{ if( !seen[$1]++ ) order[++oidx] = $1; stuff[$1] = stuff[$1] $2 "| " } END { for( i = 1; i <= oidx; i++ ) print order[i]"\t"stuff[order[i]] }' FILE_INPUT_TAB_SEPARATED | sed 's/(.*)|/\1/' 

comp100002_c0   Q9FFI3 
comp100004_c0   B9DHK3| F4J3J5 
comp100005_c0   P54150
• 1 views • link
written by EagleEye EagleEye

Thanks a million @EagleEye, I asked in stackoverflow and my question was blocked because it was 'unclear' what I was asking for. Note that to get rid of the last occurrence of '|' I used sed 's/BACK_SLASH(.*)|/\1/'. Thanks again

• 1 views • link
written by Illinu Illinu

Log in to answer this question.

More posts like this