Problem with Missing Data after Profile Alignment

[Jimmy Zheng]

Hi there,

We're encountering an issue with certain alignments losing stretches of dna after the profile alignment step. We have attached one problematic alignment, tbr1, to this post. For example, if you blast the phlawd-aligned Ophidian Holbrinkii (GI: 482713552) sequence to the one on Genbank, you'll find that the query sequence is missing 5 bases. We have looked at the sequences coming out of MAFFT and the output sequence matches the Genbank retrieved sequence exactly. However, once this sequence enters the profile aligning stage, the output sequence loses certain nucleotides and is getting broken up somehow.

We're not sure if this bug in phlawd, but it would be appreciated if we could get some insight as to how sequences are profile aligned. Is there a sequence-editing step in there? Why is it possible that some bases be deleted?

Jimmy

[mn]

Hi Jimmy,

I believe phlawd only retains conserved regions of the alignment - so perhaps this is a 5 base stretch that is absent from most of the other sequences, so this part of the alignment is not considered a "conserved block" by phlawd (and therefore, it is dropped during the profile alignment stage)?  

Best,

-Matt