[Bioperl-l] clustalw tests

Heikki Lehvaslaiho heikki@ebi.ac.uk
Fri, 11 Jan 2002 10:11:02 +0000


Jason suggested that I check clustalw 1.81 and he was right: tests work.

So the current situation with clustalw is the following:

* both 1.81 and 1.82 has serious bug in profile alingment.
  Be sure to apply the following parches is you need clustalw 
  profile functionality:

  1.81 & 1.82
  ===========
  prfalign.c. The new lines are:
  184-185:
                        alignment[nseqs1][j]=gap_pos1;
             alignment[nseqs1][j] = ENDALN;
  216-217:
                        alignment[nseqs1+nseqs2][j]=gap_pos1;
             alignment[nseqs1+nseqs2][j] = ENDALN;

  v 1.82 only
  ===========
  interface.c  Revert line 215 to
       params[i]=(char *)ckalloc(strlen(args[i])*sizeof(char));


Let's hope there will be a new bug fix release of clustalw soon.

	-Heikki



Heikki Lehvaslaiho wrote:
> 
> I am going crazy trying to understand what is going on with clustalw!
> 
> It is the profile alignment test that fails. If I rewrite the test so that
> profile_align() is passed two strings pointing to MSE files the test works.
> The current test is passed two Bio::SimpleAlign objects which forces the
> module to write both alignments into a temporary file. It is these temp
> files (or rather the first one) which clustalw program can not see and
> fails. I can not understand why not since I can change permissions of these
> files to 777 and list them before and after clustalw call:
> 
> .....|/home/heikki/tmp/PFVaKAGjRS|.......
> clustal command = /home/heikki/bin/clustalw -profile
> -profile1=/home/heikki/tmp/PFVaKAGjRS  -profile2=/home/heikki/tmp
> /PGq86fVYZO -output=gcg  -ktuple=2 -matrix=BLOSUM
> -outfile=/home/heikki/tmp/33f8uL5v4k
> 
>  CLUSTAL W (1.82) Multiple Sequence Alignments
> 
> ERROR: Could not open sequence file /home/heikki/tmp/PFVaKAGjRS)
> 
> |>ALEU_HORVU/1-362
> MAHARVLLLALAVLATAAVAVASSSSFADSNPIRPVTDRAASTLESAVLGALGRTRHALR
> FARFAVRYGKSYESAAEVRRRFRIFSESLEEVRSTN....RKGLPYRLGINRFSDMSWEE
> FQATRLG.AAQTCSATLAGNHLMRDAAA..LPETKDWREDG.IVSPVKNQAHCGSCWTFS
> TTGALEAAYTQATGKNISLSEQQLVDCAGGFNNF........GCNGGLPSQAFEYIKYNG
> GIDTEESYPYKGVNGVCHYKAENAAVQVLDSVNITLNAEDELKNAVGLVRPVSVAFQVID
> GFRQYKSGVYTSDHCGTTPDDVNHAVLAVGYGVEN.....GVPYWLIKNSWGADWGDNGY
> FKMEMGKNMCAIATCASYPVVAA
> >CATH_HUMAN/1-335
> ......MWATLPLLCAGAWLLG........VPVCGAAELSVNSLEK............FH
> FKSWMSKHRKTYS.TEEYHHRLQTFASNWRKINAHN....NGNHTFKMALNQFSDMSFAE
> IKHKYLWSEPQNCSATKS..NYLRGTGP..YPPSVDWRKKGNFVSPVKNQGACGSCWTFS
> TTGALESAIAIATGKMLSLAEQQLVDCAQDFNNY........GCQGGLPSQAFEYILYNK
> GIMGEDTYPYQGKDGYCKFQPGKAIGFVKDVANITIYDEEAMVEAVALYNPVSFAFEVTQ
> DFMMYRTGIYSSTSCHKTPDKVNHAVLAVGYGEKN.....GIPYWIVKNSWGPQWGMNGY
> FLIERGKNMCGLAACASYPIPLV
> >CYS1_DICDI/1-343
> .....MKVILLFVLAVFTVFVS...............SRGIPPEEQ............SQ
> FLEFQDKFNKKYS.HEEYLERFEIFKSNLGKIEELNLIAINHKADTKFGVNKFADLSSDE
> FKNYYLNNKEAIFTDDLPVADYLDDEFINSIPTAFDWRTRG.AVTPVKNQGQCGSCWSFS
> TTGNVEGQHFISQNKLVSLSEQNLVDCDHECMEYEGEEACDEGCNGGLQPNAYNYIIKNG
> GIQTESSYPYTAETGTQCNFNSANIGAKISNFTMIPKNETVMAGYIVSTGPLAIAADAVE
> .WQFYIGGVFDIPCN...PNSLDHGILIVGYSAKNTIFRKNMPYWIVKNSWGADWGEQGY
> IYLRRGKNTCGVSNFVSTSII..
> |------------- EXCEPTION -------------
> |------------- EXCEPTION -------------
> MSG: Clustalw call crashed: 0
> 
> STACK Bio::Tools::Run::Alignment::Clustalw::_run
> /home/heikki/src/bioperl-live/Bio/Tools/Run/Alignment/Clustalw.pm:528
> STACK Bio::Tools::Run::Alignment::Clustalw::profile_align
> /home/heikki/src/bioperl-live/Bio/Tools/Run/Alignment/Clusta
> lw.pm:485
> STACK toplevel t/c.t:90
> 
> -------------------------------------
> 
> There is a chance that clustaw is doing something strange with the -profile1
> filenames from commandline and that causes all these probelms. If you look
> carefully the error message, the file name contains an additional character.
> 
> I've asked the maintainers of clustaw to check this bug, because I can not
> trace it down. It occurs only to -profile1 and it is easy to reproduce by
> adding a character to filename 'cysprot1a.msf' which is in t/data bioperl
> directory.
> 
>         -Heikki
> 
> Jason Stajich wrote:
> >
> > I am having a hard time reproducing the TCoffee bug or Clustalw bugs.
> >
> > I have tested on Linux, solaris, and IRIX system w and w/o File::Temp
> > (running File::Temp 0.11 and 0.12) on both Clustalw and TCoffee and perl
> > 5.6.0 and have all tests running fine.
> >
> > Will try on OSF1 on sanger machines and see if I can reproduce the
> > problem there.
> >
> > --
> > Jason Stajich
> > Duke University
> > jason@cgt.mc.duke.edu
> >
> > _______________________________________________
> > Bioperl-l mailing list
> > Bioperl-l@bioperl.org
> > http://bioperl.org/mailman/listinfo/bioperl-l
> 
> --
> ______ _/      _/_____________________________________________________
>       _/      _/                      http://www.ebi.ac.uk/mutations/
>      _/  _/  _/  Heikki Lehvaslaiho          heikki@ebi.ac.uk
>     _/_/_/_/_/  EMBL Outstation, European Bioinformatics Institute
>    _/  _/  _/  Wellcome Trust Genome Campus, Hinxton
>   _/  _/  _/  Cambs. CB10 1SD, United Kingdom
>      _/      Phone: +44 (0)1223 494 644   FAX: +44 (0)1223 494 468
> ___ _/_/_/_/_/________________________________________________________
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l@bioperl.org
> http://bioperl.org/mailman/listinfo/bioperl-l

-- 
______ _/      _/_____________________________________________________
      _/      _/                      http://www.ebi.ac.uk/mutations/
     _/  _/  _/  Heikki Lehvaslaiho          heikki@ebi.ac.uk
    _/_/_/_/_/  EMBL Outstation, European Bioinformatics Institute
   _/  _/  _/  Wellcome Trust Genome Campus, Hinxton
  _/  _/  _/  Cambs. CB10 1SD, United Kingdom
     _/      Phone: +44 (0)1223 494 644   FAX: +44 (0)1223 494 468
___ _/_/_/_/_/________________________________________________________