[EMBOSS] meaning of "(Reversed)" in diffseq?

[ajb at ebi.ac.uk]

Hello Joe,

Interesting. I'm not entirely convinced it's intentional either, if
so then the documentation probably needs a little work.

To answer your question, diffseq holds the differences internally by
storing them in a feature table. The features within this table are
added using feature constructors. If some of these feature
constructors are given a start position in the sequence that is
greater than the end position then they will mark the strand of
the feature as being '-'. If the diffseq report format output sees a
strand marked as '-' then it will print (Reversed).

Owing to internal representations of differences in diffseq (for example
a start>end can be a convenient way of representing a gap)  you'll
get a '(Reversed)' feature report appearing if the second sequence has an
insertion at that point. Diffseq will generally then show the last couple
of residues of the first sequence and then the insertion in the second
sequence.

We're depleted a little by holidays here but I'll raise the issue of
whether this is intentional and/or the best way of representing/documenting
this when we're quorate again (though you may well get another response
before then depending on the order holiday email is dealt with).

HTH

Alan



> Hello,
>
> I ran  diffseq on 2  FASTA-formatted DNA sequence files on my local
> installation of EMBOSS, and I have entries in the .diffseq output file
> labeled as "(Reversed)".  Looking at both sequences in an editor, neither
> of these sequence segments is reversed (the "Sequence" string in the
> diffseq file matches the original FASTA file).  Does anyone know what is
> meant by this?  I'm analyzing the EMBOSS source code, and it looks as
> though it's related to a strand feature.  That's puzzling me, because my
> local installation has no databases installed, and it seems that this kind
> of information wouldn't be embedded in the FASTA file.  How does diffseq
> determine this strand information, and how should I interpret it?
>
> Thanks,
> Joe
> ---
>
> Joseph A. Francoeur
> Senior Software Systems Engineer
> The MITRE Corporation
> 202 Burlington Road  MS K228
> Bedford, MA  01730-1420
>
> e-mail: jfrancoe at mitre.org<mailto:jfrancoe at mitre.org>
>
>
> _______________________________________________
> EMBOSS mailing list
> EMBOSS at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/emboss
>