Rookie's plee for advice: PERL or Python, Linux or Windows?

[A1_UltiMA]

Hello kind readers,

To be able to use the majority of the command line based tools enlisted in the famous http://seqanswers.com/forums/showthread.php?t=43, which programming language would you recommend we learn: PERL or Python? and on which operating system? Are there any books that you could recommend that could bring us a little closer to your level of expertise?

We are undergrads who have been enslaved by some nice professors and forced to do their dirty Bioinformatics... supposedly we were capable enough of doing it all with a set of (free) downloadable/web-based tools, all having a graphic user interface (GUI) rather than being command line based. Without the capability of using command line tools (due to our lacking expertise) we had to resort to some quirky and dodgy comprimises. Apparently they liked the results and have come back for more, save us!

Thanks

[rdeborja]

Stick to Linux, most bioinformatics tools have a UNIX like environment in mind. I'm a Perl guy who is slowly migrating to Python. R for data visualization and some analysis is invaluable!

[themerlin]

This link:

http://korflab.ucdavis.edu/Unix_and_Perl/

has a great tutorial for basic linux and perl commands. Click 'just the documentation' to download the pdf. Do at the least the first half.

[simonandrews]

For the best chance of getting existing command line tools to work you definitely want to go with 64-bit linux.

As far as writing your own scripts to do simple processing of your data you can pretty much pick whichever scripting language you like - you should probably go with what you're more familiar with. We tend to go with Perl since we know it well, but if you're starting from scratch there probably isn't much to choose between Perl and Python. R is pretty good for visualisation, but I'm not sure I'd want to do basic data processing with it.

[A1_UltiMA]

Wow thank you all! So kind

So it seems like Linux for operating system, PERL for programming language and R for data visualisation (admittedly a new term to us, but noted nonetheless).

Thanks everyone.

[maubp]

Personally I went for Python rather than Perl, but a key factor would be what do the local experts use (people in your group or department). Having someone even a bit more experienced around to turn to for help in person is very useful - although there is plenty of stuff online.

In terms of the OS, for NGS pretty much everything works on Linux, and will probably work on the Mac (sometimes with minor tweaks), but Windows is a pain. Note that Perl, Python and R are all cross platform - I'm talking about things like assembly tools and mappers. So this is another vote for picking Linux over Windows.

Also go for a 64bit OS rather than a 32bit OS given the choice, as with NGS data you will want to be able to use more than 4GB of RAM (assuming you machine has that much, and if it doesn't you'll need it sooner rather than later).

[A1_UltiMA]

Ah thanks Maubp! Nice to read from you again.

Right, so we'll have a check with the local experts then.

Much appreciated specs advice too, but what about CPU?

Best regards.

[maubp]

Regarding machine specs, try searching this forum - there have been plenty of other threads. Also keep in mind that it will depend on what you are doing (assembling a virus is easier than a plant, mapping is easier than de novo assembly).