in reply to Re^6: Comparing adjacent lines
in thread Comparing adjacent lines

Remember I am new at Perl, so my approach is probably pretty brutish.

Input:

==> igenome.genes.onlyexons.uniq.gtf <==
chr1    unknown    exon1    815    4692    .    -    .    gene_id "LOC
+653635"; gene_name "LOC653635"; transcript_id "XR_017611.2"; tss_id "
+TSS10265";
chr1    unknown    exon2    4833    4901    .    -    .    gene_id "LO
+C653635"; gene_name "LOC653635"; transcript_id "XR_017611.2"; tss_id 
+"TSS10265";
chr1    unknown    exon3    5659    5810    .    -    .    gene_id "LO
+C653635"; gene_name "LOC653635"; transcript_id "XR_017611.2"; tss_id 
+"TSS10265";
chr1    unknown    exon4    6470    6628    .    -    .    gene_id "LO
+C653635"; gene_name "LOC653635"; transcript_id "XR_017611.2"; tss_id 
+"TSS10265";
chr1    unknown    exon5    6717    6918    .    -    .    gene_id "LO
+C653635"; gene_name "LOC653635"; transcript_id "XR_017611.2"; tss_id 
+"TSS10265";
chr1    unknown    exon6    7096    7231    .    -    .    gene_id "LO
+C653635"; gene_name "LOC653635"; transcript_id "XR_017611.2"; tss_id 
+"TSS10265";
chr1    unknown    exon7    7469    7605    .    -    .    gene_id "LO
+C653635"; gene_name "LOC653635"; transcript_id "XR_017611.2"; tss_id 
+"TSS10265";
chr1    unknown    exon8    7778    7924    .    -    .    gene_id "LO
+C653635"; gene_name "LOC653635"; transcript_id "XR_017611.2"; tss_id 
+"TSS10265";
chr1    unknown    exon9    8131    8229    .    -    .    gene_id "LO
+C653635"; gene_name "LOC653635"; transcript_id "XR_017611.2"; tss_id 
+"TSS10265";
chr1    unknown    exon10    8776    8938    .    -    .    gene_id "L
+OC653635"; gene_name "LOC653635"; transcript_id "XR_017611.2"; tss_id
+ "TSS10265";
chr1    unknown    exon11    14601    14754    .    -    .    gene_id 
+"LOC653635"; gene_name "LOC653635"; transcript_id "XR_017611.2"; tss_
+id "TSS10265";
chr1    unknown    exon12    19184    19919    .    -    .    gene_id 
+"LOC653635"; gene_name "LOC653635"; transcript_id "XR_017611.2"; tss_
+id "TSS10265";
chr1    unknown    exon13    58954    59871    .    +    .    gene_id 
+"OR4F5"; gene_name "OR4F5"; p_id "P1363"; transcript_id "NM_001005484
+.1"; tss_id "TSS13645";
chr1    unknown    exon14    77385    80096    .    +    .    gene_id 
+"LOC100132632"; gene_name "LOC100132632"; p_id "P27379"; transcript_i
+d "XM_001724183.1"; tss_id "TSS10177";
chr1    unknown    exon15    110381    110795    .    -    .    gene_i
+d "LOC643670"; gene_name "LOC643670"; transcript_id "XR_039242.1"; ts
+s_id "TSS5420";
chr1    unknown    exon16    114643    114701    .    -    .    gene_i
+d "LOC729737"; gene_name "LOC729737"; p_id "P4208"; transcript_id "XM
+_001133863.1"; tss_id "TSS9238";
chr1    unknown    exon17    118918    119086    .    -    .    gene_i
+d "LOC643670"; gene_name "LOC643670"; transcript_id "XR_039242.1"; ts
+s_id "TSS5420";
chr1    unknown    exon18    123237    130714    .    -    .    gene_i
+d "LOC643670"; gene_name "LOC643670"; transcript_id "XR_039242.1"; ts
+s_id "TSS5420";
chr1    unknown    exon19    123324    130714    .    -    .    gene_i
+d "LOC653340"; gene_name "LOC653340"; transcript_id "XR_039243.1"; ts
+s_id "TSS5420";
chr1    unknown    exon20    129326    129559    .    -    .    gene_i
+d "LOC729737"; gene_name "LOC729737"; p_id "P4208"; transcript_id "XM
+_001133863.1"; tss_id "TSS9238";
chr1    unknown    exon21    129653    129710    .    -    .    gene_i
+d "LOC729737"; gene_name "LOC729737"; p_id "P4208"; transcript_id "XM
+_001133863.1"; tss_id "TSS9238";
chr1    unknown    exon22    132810    132874    .    -    .    gene_i
+d "LOC729737"; gene_name "LOC729737"; p_id "P4208"; transcript_id "XM
+_001133863.1"; tss_id "TSS9238";
chr1    unknown    exon23    133719    134341    .    -    .    gene_i
+d "LOC729737"; gene_name "LOC729737"; p_id "P4208"; transcript_id "XM
+_001133863.1"; tss_id "TSS9238";
chr1    unknown    exon24    217633    218641    .    -    .    gene_i
+d "LOC728481"; gene_name "LOC728481"; transcript_id "XR_015292.2"; ts
+s_id "TSS13278";
chr1    unknown    exon25    313133    319752    .    +    .    gene_i
+d "LOC100132287"; gene_name "LOC100132287"; transcript_id "XR_039254.
+1"; tss_id "TSS18904";

[download]
 
#!/usr/bin/perl -w

use strict;

my $previous = '';
my $exon = 1;
while (<>) { 
  chomp;
  my(@gtf) = split; 

  chop($gtf[8] = join ":", @gtf[8,9]); 
  chop($gtf[10] = join ":", @gtf[10,11]);
  chop($gtf[12] = join ":", @gtf[12,13]);

# Restart exon# for different gene names
  if ( $gtf[9] ne $previous ) {
    $gtf[2] = 'exon1'; 
  } elsif ( $gtf[9] eq $previous ) {
    $exon =~ s/(\d+)/$1+1/e; 
  } 
  $previous = $gtf[9];
  $exon = $gtf[2];

  $gtf[8] = join "_", @gtf[8,10,12,2];
  my $bed = join "\t", @gtf[0,3,4,8];
  print "$bed\n";
}

[download]

Here's 25 lines of the output (I'll bold the lines where the problem appears and simplify the data for formatting by removing array items 10 and 12):

chr1 815 4692 gene_id:"LOC653635"_exon1

chr1 4833 4901 gene_id:"LOC653635"_exon2

chr1 5659 5810 gene_id:"LOC653635"_exon3

chr1 6470 6628 gene_id:"LOC653635"_exon4

chr1 6717 6918 gene_id:"LOC653635"_exon5

chr1 7096 7231 gene_id:"LOC653635"_exon6

chr1 7469 7605 gene_id:"LOC653635"_exon7

chr1 7778 7924 gene_id:"LOC653635"_exon8

chr1 8131 8229 gene_id:"LOC653635"_exon9

chr1 8776 8938 gene_id:"LOC653635"_exon10

chr1 14601 14754 gene_id:"LOC653635"_exon11

chr1 19184 19919 gene_id:"LOC653635"_exon12

chr1 58954 59871 gene_id:"OR4F5"_exon1

chr1 77385 80096 gene_id:"LOC100132632"_exon1

chr1 110381 110795 gene_id:"LOC643670"_exon1

chr1 114643 114701 gene_id:"LOC729737"_exon1

chr1 118918 119086 gene_id:"LOC643670"_exon1

chr1 123237 130714 gene_id:"LOC643670"_exon18

chr1 123324 130714 gene_id:"LOC653340"_exon1

chr1 129326 129559 gene_id:"LOC729737"_exon1

chr1 129653 129710 gene_id:"LOC729737"_exon21

chr1 132810 132874 gene_id:"LOC729737"_exon22

chr1 133719 134341 gene_id:"LOC729737"_exon23

chr1 217633 218641 gene_id:"LOC728481"_exon1

chr1 313133 319752 gene_id:"LOC100132287"_exon1

The bolded lines should count up from 1, for example, "exon18" should read "exon2" and exon21, exon 22, exon23 should read exon2, exon3, exon4.

Replies are listed 'Best First'.
Re^8: Comparing adjacent lines
by aaron_baugher (Curate) on Jul 16, 2012 at 19:52 UTC

    The problem here is that you're getting your $exon variable confused with $gtf[2]. You're using $exon to hold the counter which you increment on matching lines, but then you clobber it by assigning $gtf[2] to it after that, but then you don't use it after that. Also, if you're going to use a counter, you don't need to use a regex to increment the digit part of the field. Also, there's no need for an elsif when there are only two possibilities (eq and ne). 

      if ( $gtf[9] ne $previous ){
    $exon = 1;           # reset counter
  } else {
    $exon++;             # increment counter
  } 
  $gtf[2] = "exon$exon"; # put counter back in field

    [download]

    Aaron B.
    Available for small or large Perl jobs; see my home node.

[reply]
[d/l]
[select]

      Perfect!! Thank you.

      Most of those things you pointed out were an accumulation of different things I tried...and then forgot to clean up afterwards. You have been incredibly helpful, thanks again.

[reply]

In Section Seekers of Perl Wisdom