Helloo:
I just subscribed to the group and found this old message related to cDNA codon alignments
[rquiroga at fcq.unc.edu.ar](http://www.compbio.dundee.ac.uk/mailman/listinfo/jalview-discuss) wrote:
> *I am having three problems.*
>
> *One is the Translation of cDNA, which seems to not handle gaps *
> *correctly sometimes,*
> *(ATAAAG---GTTCCT) translates to (IKVP)*
> *(ATAAAA---GTTCCA) translates to (IKVP), missing the gap.*
Jim’s answer:
The translate cDNA function will exclude any gapped columns from the
translated protein alignment. Otherwise, gaps are introduced when codons
intersect but are misaligned.
e.g.
A-TAAAG
AATAA-G
becomes
-I-K
N-K-
> *Also I get a "Please select at least three bases in at least one *
> *sequence in order to perform a cDNA translation." when I try to *
> *translate my entire cDNA codon alignment to protein. (Sequences: 154 *
> *Minimum Sequence Length: 314 Maximum Sequence Length: 365 Average *
> *Length: 351). Maybe too big?*
Jim’s answer was:
My guess is that you have one or two Ns in the region you are
translating. The translate cDNA routine complains when the effective
number of *defined* bases that have been selected for translation is not
divisible by three. This is partly because it doesn't know how to use an
N or an X (or ?) in a codon.
My experience:
1. Gaps in translation are kept correctly
2. I have the same error message *"Please select at least three bases … “* when I select the whole alignment for translation and this has nothing to do with N, X or “?” in a codon.
3. This very useful option is unfortunately not working for entire sequences or large chunk of positions.
If some trick is known for fixing this problem?
Thanks
Alexey Eroshkin,
Burnham Institute for Medical Research