[kman42]

Are there any biologists running EMBOSS on OSX out there?

I know this should probably be in Software, but I wanted to cast as large a net as possible, since this is likely a pretty small group.

Anyway, I installed EMBOSS and embosslauncher.kaptn and everything seems okay, except I can't get emma to work on my local installation. I can run clustalw manually with success, but when I try to run it through emma I get the following error:

Code:
$ emma
Multiple alignment program - interface to ClustalW program
Input sequence(s): myfiles
Output sequence [cae45818.aln]: myfiles.aln
Dendogram output filename [cae45818.dnd]: myfiles.dnd
..clustalw -infile=00001370A -outfile=00001370B -align -type=protein -output=gcg -pwmatrix=blosum -pwgapopen=10.000 -pwgapext=0.100 -newtree=00001370C -matrix=blosum -gapopen=10.000 -gapext=5.000 -gapdist=8 -hgapresidues=GPSNDQEKR -maxdiv=30..

 CLUSTAL W (1.82) Multiple Sequence Alignments


Sequence type explicitly set to Protein

ERROR: Could not open sequence file 00001370Aoutfile=00001370B


No. of seqs. read = -1. No alignment!
Error: failed to open filename '00001370B'
Problem writing out EMBOSS alignment fileBus error

myfiles contains three fasta formatted sequences. This exact file works fine when run through clustalw. I'm running MacOSX 10.3.2 and I'm using the ebioinformatics.org installation of emboss, etc.

Thanks!
kman

[pimephalis]

Originally posted by kman42:

Code:
$ emma

*SNIP**

ERROR: Could not open sequence file 00001370Aoutfile=00001370B


No. of seqs. read = -1. No alignment!
Error: failed to open filename '00001370B'
Problem writing out EMBOSS alignment fileBus error

myfiles contains three fasta formatted sequences. This exact file works fine when run through clustalw. I'm running MacOSX 10.3.2 and I'm using the ebioinformatics.org installation of emboss, etc.

Thanks!
kman

I don't know this program at all, but what I'm reading above seems fairly straightforward. When you call emma, without any infile or outfile specified, it complains it can't find the file names 00001370B, which I assume is a dummy value.

If that file does indeed exist, then check the read and write permissions for the file and for the directory. It could be that they have been mis-set, and the program can't open, read, write or modify the files in question, which would cause it to fail.

[Phat Bastard]

I don't use EMBOSS but since you're just trying to run Clustalw, why don't you download Clustalw itself? You can get it here:

http://www.embl.de/~chenna/clustal/darwin/

(if you like a GUI get Clustalx, if not, stick with Clustalw)

[kman42]

Originally posted by Phat Bastard:
I don't use EMBOSS but since you're just trying to run Clustalw, why don't you download Clustalw itself? You can get it here:

http://www.embl.de/~chenna/clustal/darwin/

(if you like a GUI get Clustalx, if not, stick with Clustalw)

I have clustalw installed and it works great with the sequence files I am trying to use. I'm using EMBOSS (and spin) because they provide nice (or at least consistent) interfaces to about 185 different bioinformatics apps. I have been using clustalw separately, but I would like to get EBMOSS running properly. And in trying to get it to work, I'm hoping to learn a bit

kman