Problem With Dialign2-2 Output Sequence Names

Question

[Eric Normandeau]

Hi,

I'm using dialign2-2 to align small groups of sequences and returning fasta files containing the alignments.

This is something I have done in the past, however, I get an unexpected problem while trying to do this again today. Namely, the names of the sequences in the output alignment file are truncated to 12 characters. This is not the result I have had in the past.

Has anyone encountered such a problem? What is the solution you found?

Thanks!

(PS: My option for now is to try the more recent dialign-TX, but I have no idea if this will solve the problem.)

Answer 1

[Paulo Nuin]

IMHO I wouldn't use dialign (any version) as my alignment software of choice. The TX version, last time I checked, didn't have a good documentation and it was a big problem to make it accept the input files I had.

Try muscle or mafft and I'm pretty sure you will be happy with the result and the speed of both.

[Eric Normandeau]

Thanks for the suggestions. I'll check muscle and mafft quickly to see if they give me more interesting results.

[Eric Normandeau]

@nuin You were right about muscle. I guess this is what I should have used from the start. Do you know if muscle will try to align both sense and anti-sense sequences? Let's say one out of 10 sequences is in the wrong sense, will it be aligned properly, or should I first insure that all the sequences are in the same sense? Thanks

[Paulo Nuin]

I'm not sure about that, haven't tried something similar. Does dialign do that?

[Eric Normandeau]

Negative, dialign requires that all sequences already be in the same sense. I'll do quick tests with muscle and post the result.

[Eric Normandeau]

After testing, it seems that muscle, just like dialign, requires that input sequences be in the same sense.