Which Linux distribution is the best for Bioinformatics???

[simonandrews]

As others have said, to a great extent it doesn't matter.

Our choice is Fedora, for a few reasons:

• It's close enough to RedHat to keep the corporate types happy
• It's got the latest version of every library, so compiling packages is never an issue
• I can send people on the excellent RedHat training courses and have the knowledge be immediately applicable
• The package list has a surprising amount of bioinformatics programs in it (BioPerl, EMBOSS, Samtools etc)

To answer your original question you don't need to have an enterprise Linux distribution for bioinformatics (we even run enterprise things such as the Illumina pipeline under Fedora) - but you need to be confident in supporting yourself. If you're running a big pipeline and aren't happy sorting out problems with the OS then it can be really useful to have a number you can call where they won't tell you to RTFM.

[NextGenSeq]

Today I had to reformat all my thumb drives I was using to copy large data NGS sets.

Everytime I used Red Hat the USB drives repeatedly crashed. Reformatting the drives in Ubuntu I could copy the data even in Red Hat and now they can be read on Ubuntu, Red Hat and even Microsoft (FAT formatting).

Red Hat and Microsoft are only concerned with profits not progress, let alone user friendliness.

[sklages]

Well, I have switched from Ubuntu to Fedora, mainly because Fedora (RedHat) is heading gcc development. Compiling staden under ubuntu always was a hassle, but works like a charm under fedora :-)

But, as usual, the "best" linux distribution to use if you are not experienced is the one used in your environment, by your colleagues or friends.

cheers,
Sven

[krobison]

You'd be surprised what bioinformatics tools are in distributions. E.g., blast2 (NCBI) is available for ubuntu. I think bioperl is also (though it may not be latest version)

[ffinkernagel]

I also primarily use ubuntu.

Ubuntu is basically debian without the focus on ultra-reliability (read old versions of everything) and therefore has a broader spectrum of packeted software. It inherits debian's sane package managment, adds in an easy way for the community to publish packages (PPA) and a lot of development effort to make it 'just work.

[ymc]

I have been using Debian. I heard that Ubuntu is based on Debian but it is more popular now. What makes it better than Debian???

[Simon Anders]

It really doesn't matter. Just use a distribution you feel comfortable with. As bioinformatics tools won't be find in any package management system, you have to install them manually (which is no real problem) and this is done the same way on all Unix OSes with GNU tool chain (no matter whether it is any of the Linux distributions or Mac OS X or something more esoteric).

And, no, you don't need "Enterprise level" Linux. This is just a fancy marketing term from the RedHat people with no meaning.

Personally, I am quite happy with Ubuntu.

Simon

[NextGenSeq]

I would use Ubuntu. I hate Red Hat.

[ymc]

Is Scientific Linux any good for bioinformatics? If yes, why?

Do we need Enterprise level Linux???

What Linux distribution are you using???

Thanks in advance.