SEQanswers

SEQanswers (http://seqanswers.com/forums/index.php)
-   Bioinformatics (http://seqanswers.com/forums/forumdisplay.php?f=18)
-   -   Pairwise alignment from a consensus in query (http://seqanswers.com/forums/showthread.php?t=45305)

Antony03 07-25-2014 03:26 PM

Pairwise alignment from a consensus in query
 
Hello,
I want to look for a specific gene in a genome sequence. The gene sequence is degenerate and a colleague has suggested that I should make a multiple alignment of orthologous genes, make a consensus and look for the gene with this consensus sequence.

Do you know a software which may produce a pairwise alignment from a consensus as a query sequence ?
Thank

Brian Bushnell 07-25-2014 09:53 PM

Are you interested in protein or nucleotide alignments?

Antony03 07-26-2014 04:07 AM

Ideally, I would like to try both.

Brian Bushnell 07-27-2014 09:00 AM

Mothur can make a consensus from multiple sequences, but I'm not sure this approach will work very well; it doesn't do a very good job with highly divergent sequences. Still, worth trying. There's also PBDagCon (from PacBio) that generates a consensus from multiple sequences, though it is aimed at PacBio's error profiles rather than degenerate genes.

Antony03 07-27-2014 09:05 AM

Thank you the answer. But, how can I make a local alignment with the resulting consensus? Is fasta36 or blast can handle it?

Brian Bushnell 07-27-2014 09:15 AM

Blast should be able to, if the consensus is good enough.

kmcarr 07-27-2014 08:35 PM

PSI-BLAST was designed for this task. Provided a multiple sequence alignment of homologous sequences it creates a Position Specific Scoring Matrix (PSSM) which it then uses when searching a database for other potential homologs. This method is more sensitive than using a single sequence, even a consensus, because it weights more heavily the portions of the sequence which are more highly conserved.


All times are GMT -8. The time now is 03:32 AM.

Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2020, vBulletin Solutions, Inc.