any phylogeneticist using ubuntu?

[thermophile]

just wondering if anyone is running Paup* or arb on their ubuntu machine? I've built a few trees using fastdnaml which is massively faster than Paup on my mac, but I don't like the limited models that fastdnaml offers. I spent a few days last summer trying to get Paup and arb running on my ubuntu box but gave up cause I couldn't get them to either install or run properly. just wondering if anyone out there has any tips.

thanks

[machoo02]

I'm a grad student currently studying fish systematics and phylogenetics, and I primarily use Ubuntu. However, I don't have PAUP* running on my machine, as my lab only has a mac license for it and I can't afford the license cost of the portable version. I haven't heard of arb yet, but I will look into it. PHYLIP installs and runs great, as does PhyML and MrBayes...PhyML might be a good substitute for fastdnaml, but it also suffers from a limited choice of models (although not as much as dnaml or fastdnaml)

[thermophile]

arb is mostly a database that is also has an alignment editor and can run trees using either NJ or Phylip. I run it from X11 on my mac but it is supposed to work on linux too. it was built by a big group of microbiologist in Germany and it seems like it's mostly used by microbiologist or others looking at rRNA. but there's no reason that it couldn't work for any type of sequence data. I'd like to use MrBayes, but essentially no microbiologist use baysian trees

[takakia]

I m phylogenetist and I mostly use ubuntu. but I have a great regard for debian. Sad thing is that MrBayes segfaults in amd64 architechtrue

[frogotronic]

Hello,

An old thread, but is this still the case for MrBayes? I'm strongly considering buying PAUP4.0*beta for UNIX/LINUX. Anyone tried compiling this? Or does it come with a binary?

- CH

[mr chip]

FWIW, I'm running the standard Ubuntu package for MrBayes on 10.04 amd64 with no problems. Currently using it to infer phylogenies using hymenopteran DNA sequences from 400 odd taxa.

[tymek666]

I'm running few programs under Ubuntu:
1. MrBayes
2. ARB
3. BioEdit under Wine
4. BEAST
5. BAPS
6. SITES

Still missing program like Sequencher

[mr chip]

I've just built and run a parallel version of Mr Bayes and the improvement is making me grin like an idiot.

I tested it with primates.nex, with lset nst=6 rates=invgamma and mcmc ngen=10000 samplefreq=10. With 1 core, the job took 30 seconds, with 2 it took 16 seconds, and with 3 it took 12 seconds (these are just wallclock measurements, haven't had time to test with 4 cores yet).
Next step is to get MPI working on the network and see what effect throwing 16 cores at Mr Bayes has.

[tymek666]

I've just built and run a parallel version of Mr Bayes and the improvement is making me grin like an idiot.

Did you use this HowTo to get it work:
http://mrbayes.csit.fsu.edu/wiki/ind...ion_of_MrBayes
It says that this is for cluster, maybe because dualcore processor weren't widely available when MrBayes was released.

[mr chip]

Did you use this HowTo to get it work:
http://mrbayes.csit.fsu.edu/wiki/ind...ion_of_MrBayes
It says that this is for cluster, maybe because dualcore processor weren't widely available when MrBayes was released.

I used that as a starting point, but it's not entirely helpful for Ubuntu users. Instead of using LAM I used Open MPI, which will work on a single multicore machine and across many machines. On 10.04 desktop, I had to install mpicc, libreadline5-dev and the openmpi packages to build and run mr bayes with MPI enabled