We will be offering an R workshop December 18-20, 2019. Learn more.
Dear Pat and collaborators,
Congratulations on this nice piece of software! I have tried using Mothar for converting my sequence alignment into distances with a view to generate clusters based on those distances. However, calling dist.seqs(fasta=myfile.fas) produces files with all distances = 1. Naturally, my alignment contains similar sequences and I expected the distances to be less than 3%.
Do you know what may be going wrong?
Thanks for your help & Best wishes
--Robin 10:57, 22 July 2009 (EDT)
I think I know what the bug is - is it possible for you to put a dash at the beginning of each sequence? I suspect you somehow have sequences without any gap characters in them. I'm not sure how you pulled this off, but we've seen a bug like this in some of our testing. There should be a better fix in the next release. --Pat Schloss 14:13, 22 July 2009 (EDT)
Yes, that's it. Adding a gap column solves the problem and I get the expected distance measures. Thanks!
PS I am using mtDNA COI (which explains the absence of gaps).
In some fields pairwise sequence divergences are often computed under the K2P model. Maybe it is an idea to add this functionality?
We'll add it to the "list", thanks! --Pat Schloss 07:55, 7 August 2009 (EDT)
Would it be possible to add a square matrix form of Phylip as an output option? --Jpreem 05:35, 13 February 2010 (EST)
Yup, it's on the list --Pat Schloss 15:58, 15 February 2010 (EST)
The manual states that an alignment file can be provided in either fasta, nexus, clustal, or phylip format. Because my alignments are in nexus format, I tried something like
mothur > dist.seqs(nexus=amazon.nex) does not work because nexus is not a valid parameter.
mothur > dist.seqs(fasta=amazon.nex) works but not correctly as it finds only 1 sequence.
How should I provide Mothur with nexus files / what am i doing wrong?
--Robin 08:03, 7 July 2010 (EDT)
Oops. The manual is out of date and I'll change that. It only takes sequences in fasta format. Supporting all of the formats just got too complicated and it seemed like the other aligners (e.g. mega, clustal, etc) could output to fasta without a problem --Pat Schloss 08:08, 7 July 2010 (EDT)