Hi Kathleen,
Thanks for your post. Regarding the first error you posted, what was your work around? It
looks like the mafft invocation is looking for the link to the mafft
executable in the correct place
('satesrc-v2.2.5-2012Oct16/sate-core/bin/mafft'). After running 'python
setup.py develop' this path should be a symbolic link to
'satesrc-v2.2.5-2012Oct16/sate-core/resources/scripts/mafft'. Can you
check these paths to see if
'satesrc-v2.2.5-2012Oct16/sate-core/resources/scripts/mafft' exists and
if 'satesrc-v2.2.5-2012Oct16/sate-core/bin/mafft' is a link to it ('ls
-l' should show this)? Furthermore, if you could navigate to 'satesrc-v2.2.5-2012Oct16/sate-core/bin' and run 'ls -l *' to check if the setup script successfully created links to the various tools, that might be helpful. You should get something like the following, but with different paths:
####
lrwxrwxrwx 1 jamie jamie 56 Jan 2 19:45 clustalw2 -> /home/jamie/software/dev/sate/sate-tools-linux/clustalw2
lrwxrwxrwx 1 jamie jamie 58 Jan 2 19:45 fakealigner -> /home/jamie/software/dev/sate/sate-tools-linux/fakealigner
lrwxrwxrwx 1 jamie jamie 55 Jan 2 19:45 faketree -> /home/jamie/software/dev/sate/sate-tools-linux/faketree
lrwxrwxrwx 1 jamie jamie 55 Jan 2 19:45 fasttree -> /home/jamie/software/dev/sate/sate-tools-linux/fasttree
lrwxrwxrwx 1 jamie jamie 57 Jan 2 19:45 fasttreeMP -> /home/jamie/software/dev/sate/sate-tools-linux/fasttreeMP
lrwxrwxrwx 1 jamie jamie 63 Jan 2 19:45 mafft -> /home/jamie/software/dev/sate/sate-core/resources/scripts/mafft
lrwxrwxrwx 1 jamie jamie 53 Jan 2 19:45 muscle -> /home/jamie/software/dev/sate/sate-tools-linux/muscle
lrwxrwxrwx 1 jamie jamie 55 Jan 2 19:45 opal.jar -> /home/jamie/software/dev/sate/sate-tools-linux/opal.jar
lrwxrwxrwx 1 jamie jamie 57 Jan 2 19:45 padaligner -> /home/jamie/software/dev/sate/sate-tools-linux/padaligner
lrwxrwxrwx 1 jamie jamie 52 Jan 2 19:45 prank -> /home/jamie/software/dev/sate/sate-tools-linux/prank
lrwxrwxrwx 1 jamie jamie 55 Jan 2 19:45 randtree -> /home/jamie/software/dev/sate/sate-tools-linux/randtree
lrwxrwxrwx 1 jamie jamie 52 Jan 2 19:45 raxml -> /home/jamie/software/dev/sate/sate-tools-linux/raxml
lrwxrwxrwx 1 jamie jamie 53 Jan 2 19:45 raxmlp -> /home/jamie/software/dev/sate/sate-tools-linux/raxmlp
real_bin:
total 16
lrwxrwxrwx 1 jamie jamie 66 Jan 2 19:45 disttbfast -> /home/jamie/software/dev/sate/sate-tools-linux/real_bin/disttbfast
lrwxrwxrwx 1 jamie jamie 63 Jan 2 19:45 dvtditr -> /home/jamie/software/dev/sate/sate-tools-linux/real_bin/dvtditr
lrwxrwxrwx 1 jamie jamie 70 Jan 2 19:45 pairlocalalign -> /home/jamie/software/dev/sate/sate-tools-linux/real_bin/pairlocalalign
lrwxrwxrwx 1 jamie jamie 62 Jan 2 19:45 tbfast -> /home/jamie/software/dev/sate/sate-tools-linux/real_bin/tbfast
####
If you have run setup.py without errors and do not see links here, let me know. If you do see links, check to see if the link targets (the path after '->') exist.
Assuming you find that the links and their targets do exist, the second error you report is likely due to the fasttree binaries that are distributed with the SATe source (in the 'sate-tools-linux' directory). They are both dynamically linked 32-bit builds (I've recently replaced them with statically linked 32-bit builds on github, but after the 2.2.5 release). We distribute SATe with 32-bit builds to try and maximize compatibility, but there can be problems with the 32-bit multi-threaded version ('fasttreeMP') on 64-bit linux systems. It will run fine on one processor, but often does not when using multiple processors.
The best solution for this problem is to download the 64-bit multi-threaded executable from
http://www.microbesonline.org/fasttree/#Install (or download the source from here and compile the multi-threaded version on your cluster). Then, rename the executable as "fasttreeMP" and replace the existing executable in the "sate-tools-linux" directory. Let me know if you have questions about how to do this or if it does not fix the problem.
Sorry for the long response, but hopefully some of it is helpful! I have used SATe on Linux-based clusters, so I'm confident that we can get you up and running!
Jamie