Open cjfields opened 8 years ago
cjfields
commented
8 years ago Original Redmine Comment Author Name: Chris Fields Original Date: 2011-09-20T14:13:29Z
This seems to be a problem with CUTG output and not bioperl. I have notified them of the issue, let’s see if they respond. Otherwise, there isn’t much we can do from our end beyond switch output (which will require rewriting the parser from scratch, not the best option).
cjfields
commented
8 years ago Original Redmine Comment Author Name: Keith Robison Original Date: 2011-09-20T18:01:47Z
Yeah, it’s clearly an upstream problem. But, these are easy to compute — some code I wrote after posting this last night to fix this (where $cut is the usage table)
foreach my $three(keys %{$cut->{’_table’}}) { my $sum=0; foreach my $hash(values %{$cut->{’_table’}->{$three}}) { $sum+=$hash->{’abs_count’}; } foreach my $hash(values %{$cut->{’_table’}->{$three}}) { $hash->{’rel_freq’}=$hash->{’abs_count’}/$sum; } }
cjfields
commented
8 years ago Original Redmine Comment Author Name: Keith Robison Original Date: 2011-09-20T18:07:54Z
What other thought — the code above is not rounding anything the way the current tables (and IO) do, so there is a risk of slightly different behavior between loading a file from disk & slurping one from the server if the code above is put in place
Author Name: Keith Robison (Keith Robison) Original Redmine Issue: 3292, https://redmine.open-bio.org/issues/3292 Original Date: 2011-09-20 Original Assignee: Bioperl Guts
The code below on a fresh install of BioPerl will print 0 for the relative frequency of ATG for Bacillus subtilis (or apparently any other organism). The Kasuza site is apparently not computing this any more & returning ‘0.00’ in all cases.
!/usr/bin/perl
use strict; use Bio::CodonUsage::Table; use Bio::DB::CUTG;
my $db=Bio::DB::CUTG->new(-sp =>‘Bacillus subtilis [’, -gc => 1); my $cut=$db->get_request(); print “Look! No Met!\t”,$cut->codon_rel_frequency(‘ATG’),“\n”;