[simonandrews]

No doubt that the best option is ubuntu linux.

Because....?

[TonyBrooks]

There is no best option specifically for bioinformatics. As mentioned previously, software can be compiled on any Linux system. More important is to find one you are comfortable with for the other functions that you may find useful.
Ubuntu is probably the most user friendly Linux distro. Most things work out of the box regardless of hardware and pretty much all non bioinformatics software is included in the repositories for easy install (OpenOffice, Filezilla etc). I just think there's less of a learning curve with Ubuntu for those who are comfortable with Windows or OS X (which constitute almost all of our collaborators).

[ulz_peter]

I personally prefer Ubuntu as well (just found out that bwa and samtools is available in Ubuntu repositories), just because it is easy and getting help is facilitated by the great amunt of people using it. However I had trouble installing the Roche software.
Does anyone use the Bioinformatic Ubuntu distribution biolinux?:

http://nebc.nerc.ac.uk/tools/bio-linux/bio-linux-6.0

For file sharing with PIs, I think OpenOffice is sufficient as it can handle all the Ms office data formats.

[TonyBrooks]

Yes, the 454software is a bit of a pain to get installed and Roche won't support anything other than Redhat, Fedora or CentOS. There are posts on SeqAnswers on how other people have managed to install on Ubuntu though.

[ulz_peter]

Yes, the 454software is a bit of a pain to get installed and Roche won't support anything other than Redhat, Fedora or CentOS. There are posts on SeqAnswers on how other people have managed to install on Ubuntu though.

I actually ended up changing one of my computer's operating system to Fedora, before the solution was published on Seqanswers. As it runs Gnome as well the changes are minor, mostly only affecting directory structure...and bwa and samtools are in the Fedora repositories as well

[andreitudor]

I use Bio-Linux 6. It is based on ubuntu, but is developped specifically for bioinformatics by NEBC. It has a lot of software already installed. Of course you have to update some of the tools, but everything works out of the box. You just install it and you have some 300 bioinfomatics programs and scripts.

[Daniel Fernandez]

1) Is Scientific Linux any good for bioinformatics? If yes, why?
No, a good bioinformatician is who know how to analyze the sequences and interpreted the results, not who know Linux, Linux is only a good system to work. And depends of the area, for example I know a lot of people workking in phylogeny and they don't use Linux, they don't need, maybe other lab in phylogeny used, but not all, and both are very good in that, in my case I work in functional genomics, expression, so I need to know Linux. And I know other people that know a lot of Linux but they are not good bioinformatician.

2) Do we need Enterprise level Linux???
No, you don't need enterprise level, I explain that above.

3) What Linux distribution are you using???
Well, if you decide work in Linux, you can choose the distro that you prefer, in which you feel most comfortable. Don't confuse the most popular, with the best or the most stable, the most populra is Ubuntu and is very good for work, if you are a beggining I think Ubuntu is the best option, but Fedora is the most stable and Fedora is 6 months more current than Ubuntu, about the kernel, usual softwares (not bioinfo softwares). Both are the most uses you choose any that you like. In the beggining I used Ubuntu (deb packages) now I use Fedora (rpm packages) but most bioinformatic tools are in .tar.gz or .zip for any distro in Lunix.
If you are a beginner you can use BioLinux, Open Discovery or other special softwares for bioinformaticians.

The first question must be, what do you do in bioinformatics?

[sklages]

1) Is Scientific Linux any good for bioinformatics? If yes, why?
No, a good bioinformatician is who know how to analyze the sequences and interpreted the results, not who know Linux, Linux is only a good system to work. And depends of the area, for example I know a lot of people workking in phylogeny and they don't use Linux, they don't need, maybe other lab in phylogeny used, but not all, and both are very good in that, in my case I work in functional genomics, expression, so I need to know Linux. And I know other people that know a lot of Linux but they are not good bioinformatician.

hmm, you are probably right if the linux system is centrally administered.
If you run it on your "personal" machine or your labtop you do need to know a bit of linux. Having happily installed your own linux on your notebook, you are probably happy, if you find your software of question in the distro's repos without having to think about how to compile that stuff.
And finally you are right, that a bioinf should know how to analyze his/her data/sequences ;-)

2) Do we need Enterprise level Linux???
No, you don't need enterprise level, I explain that above.

Enterprise level? Voluntary? No. But if you are dependent on commercial software it may happen that this piece of software require sth. like CentOS (RedHat) or Suse (SEL) to get support.

3) What Linux distribution are you using???
Well, if you decide work in Linux, you can choose the distro that you prefer, in which you feel most comfortable. Don't confuse the most popular, with the best or the most stable, the most populra is Ubuntu and is very good for work, if you are a beggining I think Ubuntu is the best option, but Fedora is the most stable and Fedora is 6 months more current than Ubuntu, about the kernel, usual softwares (not bioinfo softwares). Both are the most uses you choose any that you like. In the beggining I used Ubuntu (deb packages) now I use Fedora (rpm packages) but most bioinformatic tools are in .tar.gz or .zip for any distro in Lunix.
If you are a beginner you can use BioLinux, Open Discovery or other special softwares for bioinformaticians.

The first question must be, what do you do in bioinformatics?

Here I totally agree. Use whatever linux you feel comfortable with or what your environment is using (if so). On the other hand, for linux beginners (with bioinformatic focus) Ubuntu is probably NOT the best choice. I have worked with quite a few distros, but things are sometimes harder to get running on Ubuntu than e.g on a RedHat-based distro like Fedora or even CentOS. E.g. Ubuntu uses 'dash' as the system shell; though dash is POSIX compliant and should have no problems with sh-shellscripts, it turned out that quite a few sh-shellscripts are NOT POSIX compliant and as a consequence installation of program ABC fails. This is not a problem for a single shellscript but may be annoying for larger packages containing more shell scripts (e.g. Roche's 454 software suite).

This is even worse when compiling packages on your own; compiling Roche's gsMapper on Ubuntu is really a hard job, whereas on Fedora it compiles like a charm. This is mainly due to available gcc versions (in the repos) and availabilty of the dev packages ..

just my 2p :-)

Sven

[A_Morozov]

Are those specialised bioinformatics distros any good? I mean, you can print something like "sudo apt-get install bioperl" (which by the way drags lots of basic tools like aligners, blast and some more as dependencies) whenever you like. And you do not litter your bin with random stuff that does God knows what and is definitely unnecessary for you.