Mercurial > repos > earlhaminst > blast_parser

comparison blast_parser.pl @ 0:87eda806422d draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload for repository https://github.com/TGAC/earlham-galaxytools/tree/master/tools/blast_parser commit 75c6b4d9bd23cdd5f8e5626b1b01f2abba32c274-dirty
author earlhaminst
date Mon, 12 Dec 2016 07:13:57 -0500
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:87eda806422d
1 #!/usr/bin/perl
2 use strict;
3 use warnings;
4 use List::Util qw(min max);
5
6 # A simple Perl parser to convert a BLAST 12-column or 24-column output into a
7 # 3-column input for hcluster_hg (id1, id2, weight):
8 # parse_blast.pl <file>
9
10 use constant LOG_E_10 => log(10);
11
12 my $file1 = $ARGV[0];
13 open my $fh1, '<', $file1;
14
15 while (my $line = <$fh1>) {
16 my @row = split(/\t/, $line);
17
18 if ($row[0] eq $row[1]) {
19 # ignore self matching hits
20 } else {
21 # Convert evalue to an integer weight with max 100
22 my $weight = 100;
23
24 #if the evalue is 0, leave weight at 100
25 if ($row[10] != 0 && $row[10] != 0.0) {
26 $weight = min(100, positive_round(-1 * log10($row[10])));
27 }
28 print"$row[0]\t$row[1]\t$weight\n";
29 }
30 }
31 close $fh1;
32
33 # Calculate logarithm to base 10 of a number
34 sub log10 {
35 my $n = shift;
36 return log($n) / LOG_E_10;
37 }
38
39 # Round a positive float to the nearest integer
40 sub positive_round{
41 my $n = shift;
42 return int($n + 0.5);
43 }