[phylocom-user] Phylocom input files
yjlim8 at gmail.com
Fri May 28 11:51:44 PDT 2010
Actually, my phylogeny wasn't in nwk format. I fixed the problem and
got the results. And I forgot to type ./ in front of phylocom phydist
command. Thank you.
I have another problem. My NRI results from two different phylogenies
from the same alignment gave me constrasting results. Is this common?
A parsimony tree was created by a bootstrap method with fast
stepwise-addition with 1,000 replicates in PAUP. There was a
significant difference among my samples from this tree. The other was
created by a likelihood analysis in RAxML with 378 replicates with a
general time reversible model. This tree gave me NRI results that are
similar among the samples.
Any suggestions/comments would be appreciated.
On Mon, May 24, 2010 at 1:47 PM, Lim <yjlim8 at gmail.com> wrote:
> Hi Steve,
> Thanks for the reply. I have only the numbers for the short branch
> lengths (no non-numeric characters) in the attached file, likeli.nwk.
> And I sorted the sample names in my sample file, likeli.txt
> (attached). However, I still get the error message.
> And when I type "phylocom phydist -f likeli.nwk", I get "command not found."
> Could you have a look at the files again? Thanks!
> On Thu, May 20, 2010 at 3:25 PM, Steven Kembel <steve.kembel at gmail.com> wrote:
>> Hi Seara,
>> Your phylogeny has scientific notation for some of the really short branch lengths (i.e. "1.44515208833423E-6"). Phylocom doesn't deal well with non-numeric characters in branch lengths. I'd suggest either collapsing very short branches to polytomies or reformatting the branch lengths. You can do this in R using the ape package function di2multi (to collapse short branches to polytomies). And the write.tree function in ape allows you to specific the number of digits used for printing branch lengths. Read your tree into R with read.tree, collapse the polytomies if you want, write it out again with write.tree and the digits parameter set to a value that gets rid of the non-numeric characters in your branch lengths.
>> Your sample file needs to be sorted by sample name. Please have a look at the Phylocom manual for a detailed explanation of how the sample file should be formatted.
>> The branch length issue is coming up a lot lately, I will add the solution above to the manual in future versions. For others having trouble loading files into Phylocom, a troubleshooting tip is to first try loading just the phylogeny. You could do this by typing "phylocom phydist -f phyloFileName", since this function loads just the phylogeny. If this fails, there is a problem with your tree. Often branch length formatting is the issue if the tree otherwise appears to be a valid Newick string. If phydist works but comstruct or other functions do not work, the problem is with your sample or traits file. There are a few functions in the picante package for R (writesample, writetraits) that can be used to easily generate properly formatted sample and traits files for use by Phylocom.
>> On May 19, 2010, at 9:25 PM, Lim wrote:
>>> I'd greatly appreciate if someone can help me figure out what's wrong
>>> with my input files!
>>> The sample data is in txt format (attached), and the final line has a
>>> return character. The newick tree file has a semicolon at the end,
>>> and the taxa names do not begin with a number. However, I'm getting
>>> the error message:
>>> File `likelihood.txt' has unix newlines
>>> Bus error
>>> Seara Lim
>>> MSc student
>>> University of British Columbia
>>> phylocom-user mailing list
>>> phylocom-user at lists.phylodiversity.net
More information about the phylocom-user