从文件中提取特定数据并将其写入另一个文件

Question

I tagged python and perl in this only because that's what I've used thus far. If anyone knows a better way to go about this I'd certainly be willing to try it out. Anyway, my problem:

I need to create an input file for a gene prediction program that follows the following format:

seq1 5 15
seq1 20 34

seq2 50 48
seq2 45 36

seq3 17 20

Where seq# is the geneID and the numbers to the right are the positions of exons within an open reading frame. Now I have this information, in a .gff3 file that has a lot of other information. I can open this with excel and easily delete the columns with non-relevant data. Here's how it's arranged now:

PITG_00002  .   gene    2   397 .   +   .   ID=g.1;Name=ORF%
PITG_00002  .   mRNA    2   397 .   +   .   ID=m.1;
**PITG_00002**  .   exon    **2 397**   .   +   .   ID=m.1.exon1;
PITG_00002  .   CDS 2   397 .   +   .   ID=cds.m.1;

PITG_00004  .   gene    1   1275    .   +   .   ID=g.3;Name=ORF%20g
PITG_00004  .   mRNA    1   1275    .   +   .   ID=m.3;
**PITG_00004**  .   exon    **1 1275**  .   +   .   ID=m.3.exon1;P
PITG_00004  .   CDS 1   1275    .   +   .   ID=cds.m.3;P

PITG_00004  .   gene    1397    1969    .   +   .   ID=g.4;Name=
PITG_00004  .   mRNA    1397    1969    .   +   .   ID=m.4;
**PITG_00004**  .   exon    **1397  1969**  .   +   .   ID=m.4.exon1;
PITG_00004  .   CDS 1397    1969    .   +   .   ID=cds.m.4;

---

So I need only the data that is in bold. For example,

PITG_0002 2 397

PITG_00004 1 1275
PITG_00004 1397 1969

Any help you could give would be greatly appreciated, thanks!

Edit: Well I messed up the formatting. Anything that is between the **'s is what I need lol.

Answer 1

In Unix:

grep <file.gff3 " exon " |
    sed "s/^\([^ ]+\) +[.] +exon +\([0-9]+\) \([0-9]+\).*$/\1 \2 \3/"

Answer 2

For pedestrians:

(this is Python)

with open(data_file) as f:
    for line in f:
        tokens = line.split()
        if len(tokens) > 3 and tokens[2] == 'exon':
            print tokens[0], tokens[3], tokens[4]

which prints

PITG_00002 2 397
PITG_00004 1 1275
PITG_00004 1397 1969

Answer 3

It looks like your data is tab-separated.

This Perl program will print columns 1, 4 and 5 from all records that have exon in the third column. You need to change the file name in the open statement to your actual file name.

use strict;
use warnings;

open my $fh, '<', 'genes.gff3' or die $!;

while (<$fh>) {
  chomp;
  my @fields = split /\t/;
  next unless @fields >= 5 and $fields[2] eq 'exon';
  print join("\t", @fields[0,3,4]), "\n";
}

output

PITG_00002  2 397
PITG_00004  1 1275
PITG_00004  1397  1969

Answer 4

Here's a Perl script option perl scriptName.pl file.gff3 :

use strict;
use warnings;

while (<>) {
    print "@{ [ (split)[ 0, 3, 4 ] ] }\n" if /exon/;
}

Output:

PITG_00002 2 397
PITG_00004 1 1275
PITG_00004 1397 1969

Or you could just do the following:

perl -n -e 'print "@{ [ (split)[ 0, 3, 4 ] ] }\n" if /exon/' file.gff3

To save the data to a file:

use strict;
use warnings;

open my $inFH,  '<',  'file.gff3' or die $!;
open my $outFH, '>>', 'data.txt'  or die $!;

while (<$inFH>) {
    print $outFH "@{ [ (split)[ 0, 3, 4 ] ] }\n" if /exon/;
}