> > 6. Edit->Translate, turn on Align Translation, and hit OK.
>> Here's where I had the problem: I got a single line of translation,
> all in
> the reading frame of the first CDS. If I select each CDS separately and
> translate it (remembering to select align translation each time -- why
> isn't this the default?), it works.
>Here I must disagree with you to some extent. Seqlab joins the selected
CDS's before it translates and aligns them. Thus it properly takes care of
codons splitted by introns when you select all exons simultaneously.
If you select separately the exons coding for one single protein, then you
may get quite a wrong translated sequence.
Let's take the example of L29190 once again. If I do what Steve suggested
I get the correct protein sequence aligned to corresponding exons :
...ATGGGGGCGATG (...) ACCCGCGCTG//GTGAG (...) CCCAG//GCTCACACTCGCTGA...
...M~~G~~A~~M~~ (...) T~~R~~A~~G//~~~~~ (...) ~~~~~//~~S~~H~~S~~L~~R...
If I select each exon separately though, I get :
...ATGGGGGCGATG (...) ACCCGCGCTG//GTGAG (...) CCCAG//GCTCACACTCGCTGA...
...M~~G~~A~~M~~ (...) T~~R~~A~~X
...~~~~~~~~~~~~ (...) ~~~~~~~~~~//~~~~~ (...) ~~~~~//A~~H~~T~~R~~*~~
...which, obviously, is not what we want...
(// to show splice junctions are mine)
Of course, if more than one protein is coded by your genomic sequence,
then you'd better select each one separately to avoid the problem you
encountered. But, again, all exons for every single protein should be
selected simultaneously.
In any case though, you should check that the first selected exon is
translated in the correct reading frame. Which may not be the case
if it is not exon1... In this case, I'm afraid you'll need to select
the first exon at the sequence level to exclude the first one or two
bases.
Cheers,
Fred
