Genome Alignment considering ORFs (AA)

[Oxygen81]

Hello everybody,

I am searching for an alignment tool which generates multiple alignments. But a specific feature should be that it considers ORFs. It should align ORFs on the basis of the amino acids and intergenic reagions on nucleotide level. Has anyone heard about something like that?

Thank you very much!

[nickloman]

Not exactly sure what you mean but you might find promer (part of MUMMER) useful which will do pair-wise genome alignment with the conceptual 6-frame translation of the nucleotide sequences.

[nickloman]

BTW you might want to consider whether you really mean open reading frames (ORF) or coding sequence (CDS).

[nickloman]

(http://roughguidetoevolution.blogspo...-sequence.html)

[Oxygen81]

Hi Nickloman,

you are right I mean CDs. The problem with Mummer is the pairwise comparison but I need a multiple alignment. Thanks anyway!

[nickloman]

You could use Mauve and then write a script to parse the multiple alignment to extract the relevant regions for further amino acid alignment...

[bckirkup]

Context based alignment

Nick;
This is a capability that many people have built/scripted, I think, on their own - but I haven't seen a good multiple sequence alignment pipeline that preserves, for example, the codons of CDs, the known regulatory elements, and small RNAs by secondary structure, returning the primary sequence aligned at the bp level.

Do you know of one I haven't seen?

[nickloman]

Nope, as you say many people have scripted something approximating this requirement (including me) but I am not aware of a package for this.

[bckirkup]

Nick;
Maybe it is time we distributed something for it. It would ideally be able to handle progressive alignments of regions 3-100 kb, taking a bush (guide tree), raw sequences, and one or more annotated sequences as the main arguments, outputting the alignment; perhaps also pipelines for gene trees and transferred annotations.

My original vision, a long time ago, was that it would take raw sanger reads, as well as meta-information about the reads, so that it could improve base calls, assemble the reads, align them and build trees in an iterative fashion - useful if one were doing amplicons from isolates, for, say functional genes or MLST targets. Now, with the proliferation of sequencing technologies, it is not as relevant to focus on Sanger.

[Oxygen81]

Hi,

thank you for your answers! So if you plan to publish something like that let me know