3
$\begingroup$

I am converting outputs in stockholm format to clustalw using

Bio::AlignIO->new(-file => "$ARGV[0]", -format => "stockholm"); and

Bio::AlignIO ->new (-format =>'clustalw');

I am also using

my $new=$aln->remove_columns(['all_gaps_columns'],'.');

and got an output like

head1  ----------------.GG-.-gggaguggugugguuacgaaugUGGCCUCUGC-----A

head2  GGGGGUGUAGCUCAGU.GGU.A......................GAGCGGAUGCUUUGCA

What is the meaning of dots and dashes? Is this natural in a clustalw output? Am I getting a bug?

I was unable to find the sequences I used in the original message but these two pose a similar pattern. These are RNA sequences, including pseudogenes, so there are lots of palindromic sequences.

# STOCKHOLM 1.0
#=GF AU Infernal 1.1.1

head1          --------UGGAGAAU.G--.-ugggcguc.gaucccacUUCCUCUCGCAUGCUAA..............................................................GCGAGC-gcucuaccacuugagcuaauucccc-..-------------..................................
 #=GR head1 PP ........89999988.4....6789****.********999**************..............................................................******.7999999866665555555554444..................................................

 head2         --------UGGAGAAU.---.-gcgggcaucgaucccgcUUCCUCUCGCAUGCUAA..............................................................GCGAGCG.........................-..-------------cucuaccauu........................
 #=GR head2  PP ........89999887......46789************9999*************..............................................................*****86.........................................4555666666........................


 CLUSTAL W (1.81) multiple sequence alignment

 head1  --------UGGAGAAU.G--.-ugggcguc.gaucccacUUCCUCUCGCAUGCUAA....
 head2  --------UGGAGAAU.---.-gcgggcaucgaucccgcUUCCUCUCGCAUGCUAA....
$\endgroup$
  • $\begingroup$ Dash is used for gaps and dots are used for semi-conserved substitutions (similar residues). But this looks weird. Are you sure this is how you got the output because the dots are present below the alignment (next line)? $\endgroup$ – WYSIWYG Feb 7 '15 at 6:23
  • $\begingroup$ Sorry for putting this in the Answer box, but I don't have enough rep to comment yet... Can you please post the contents of your stockholm file as well? Also, just to be 100% clear, you are using BioPerl for your conversion, yes? $\endgroup$ – Steve Bond Feb 7 '15 at 13:32
  • 1
    $\begingroup$ This looks like you are running clustalw with bioperl. I have answered assuming the original alignment is wrong. Could you post your original alignment, and what versions of these are you using? $\endgroup$ – James Feb 7 '15 at 14:45
  • $\begingroup$ Please show us the sequences you are attempting to align. They look like low complexity regions which are very hard to align but we can't judge without seeing the originals. $\endgroup$ – terdon Feb 7 '15 at 15:28
  • $\begingroup$ Yes, I am using Bioperl. I uploaded a sample of the original. $\endgroup$ – je_b Feb 7 '15 at 21:04
2
$\begingroup$

As WYSIWYG touches on in the comments, usually dashes mean that there are gaps in the alignment as the result of an indel event. Dots are used to show specifically point mutations where the biochemical properties are mostly conserved, but the residue has changed.


From original question, not relevant to question as it stands.

You are right in that this isn't a proper alignment, however I'm not sure it is a bug. What you have appears to be the alignment of a nucleotide sequence that has gone horribly, horribly wrong.

As a first guess without any more details I would say it doesn't look like you have installed clustalw correctly or pointed your bioperl to it. From the documentation here are common problems:

  1. Make sure the clustalw executable is in your path so that which clustalw returns a clustalw executable on your system.

  2. Define an environmental variable CLUSTALDIR which is a directory which contains the 'clustalw' application: In bash:

    export CLUSTALDIR=/home/username/clustalw1.8

    In csh/tcsh:

    setenv CLUSTALDIR /home/username/clustalw1.8

  3. Include a definition of an environmental variable CLUSTALDIR in every script that will use this Clustalw wrapper module, e.g.:

    BEGIN { $ENV{CLUSTALDIR} = '/home/username/clustalw1.8/' } use Bio::Tools::Run::Alignment::Clustalw;

Unless you are doing massive amounts of sequence alignment, the clustalW2 webserver is quicker, more sensitive and way easier to use.

| improve this answer | |
$\endgroup$
1
$\begingroup$

Based on the edit you made to your op, the dots are present in your initial stockholm alignment, and are simply being copied into the new clustal output.

What is the source of your alignment to begin with? I'm assuming you're selecting individual rows from a larger alignment, given the odd distribution of gaps. This could give you a clue as to the actual significance of the dots (dashes are pretty much always gaps).

Also, just fyi, if you want a command line tool to do the conversion for you, I have a python3 module called SeqBuddy that will work just fine. You'll need BioPython as a dependency though.

$: python3 SeqBuddy.py input.stockholm -sf clustal > output.clustal
| improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.