Entering edit mode
8.0 years ago
nchuang
▴
260
Quick question. A Multiple Sequence alignment (MSA) uses dynamic programming to create alignments. Would it still be able to give a good global pairwise alignment? I tried it with TCoffee and result looked good for a query sequence with large indel and it was fast. I am just trying to get a collection of sequences aligned to a reference sequence (preferable output in fasta format). I was trying Python Pairwise2 but apparently it can't do large sequences (>3500bases) without being extremely slow.
Pairwise alignment methods are built into t-coffee. See the manual.
ok thanks I was reading their manual and it didn't specifically say so. I think Mafft stated it uses Needleman for global alignments.
Biopython's
pairwise2has been improved regarding performance. The updated version is not merged with the main branch but you may downloadpairwise2.pyandcpairwise2module.cfrom this pull request. The C file must be compiled.I had a lot of trouble with pairwise2. It was taking forever and I think I read somewhere that it is just too slow? I don't know if mine was the updated version.
I ended up just doing Needleman on all of them to the same reference and then used biopython to arrange them to the same positions as the reference in a numpy matrix.