I understand that Computational Biology is a broad field and the answer may depend on the specific work one has to develop but let's imagine we have to recommend an OS to somebody with a recent degree in Health Sciences with only user-level knowledge of computers but who wants to focus his career on Computational Biology.
Thanks.
Thank you all for your answers.
I personally think that MAC OS X is a good option to have a "ready-to-go" powerful system based on UNIX but for someone who wants to start a career in computational biology (so, we are talking about a long term investment) it's worth to learn Linux.
I would recommend Linux over Windows/Mac because its free and open source software (FOSS) having lots of vibrant features. Some of the features are almost free from any virus attack, less crash rate, many compilers comes in free with the OS, good for multithreading and time sharing system, can be used as stand-alone as well as a server.
In context of Computation Biology and Bioinformatics, almost all the software tools are available on Linux platform.
it depend what are you want to do. I'm using 4 OS'a:
1. Windows - a lot of tools work ONLY in Windows (very often don't work in emulators or VM or are not so effective - for example some GUI on .NET). I use very often VBA for excel (is very easy) and use it under Win is very easy or programming using .NET for C++ .
2. Mac OS - for molecular modeling and visualisation
3. different distributions of Linux for very long calculations (for example molecular dynamics). Linux is the most "flexible" system, you can eaven compile the system with the configuration you want to have (several years ago I used the program - UHBD where I had to compile the system to set any changes in the program configuration - it is possible ONLY under Linux). Using linux you can utilize the hardwere most effective for your specific program.
for not very specific calculations and not to have any problems with the machine I recomend Mac OS.
I use Bio-linux 7 (http://nebc.nerc.ac.uk/nebc/tools/bio-linux) as my main system
I recommend Linux. Most of the Bioinformatics tools are compatible with Linux OS.
Probably linux is a better choice in general since it helps with long calculations and it is easier to setup in server-cluster configurations
There is no simple answer, even with the extra conditions you have formulated. As Jan has pointed, the variety of tools is such that no single platform can harbor them all. An expert user would solve the problem easier, but beginners depend on 3rd party applications to do the job. If you know your toolbox, say, Matlab, than just go for the easiest most comfortable platform on which it runs. With no specific area. If you intend to have a variety of tools and problems to address and really limited to user-level interaction with your computer I would recommend to settle on a good Mac with a nice big screen. It will give you direct uncomplicated access to the largest part of Bioinformatics software. It's easy and comfortable, but runs on top of OS X, which is kind of Unix. Most Linux applications will port seamlessly. Mac is not an obscure minority platform it used to be over decade ago. Most proprietary software packages associated primarily with Wintel platform have been ported and available for Mac as well. There will be still some pieces of software out of your grasp. Some free Linux programs you download will refuse to compile, some will work incorrectly, others will still be only for PC running Windows. I don't know if it makes you feel better, but most programs that don't port from Linux to Mac are poorly designed, they are shaky and tricky on any platform. You will be past your beginner stage before you can use them any way. Choosing Mac will not solve all the problems, it will put more software within your reach than any other choice.
As for myself - I don't use Mac at all. I sit on Win7 and have Cygwin running locally plus Xming with open Xterm on one ore more of Linux (or other Unix) high-performance computing clusters. Don't follow me right away - I'm not a beginner.
Linux (Fedora-Ubuntu) is the best also for me.. There are several ways to make windows stuff working in linux... the ultimate is to install a virtualbox and making windows working within linux system installing everything you can not making work on linux.. My experience is that at the end you will find your way to perform almost all of your analyses on linux
Ultimately it depends entirely on what you want to do. If you don't want the primary OS to be command line-driven then apple OS offers the benefits of having a GUI for the OS but easy access to linux command line interface through the terminal. Java-driven software allows access to most open-source bioinformatics software so that is not so much platform dependent.
You can installed dual OS (Windows and Linux). But Linux most great tool and database of the Bioinformatics software.
If you want to learn how to use use a linux distribution (as Fedora, RedHat), the easier way is to install a virtual box that will allow to emulate a LINUX OS.
To my point of view, if you want to dig how the software is running and enter in the kingdom of bioinformatics, Linux OS are the best. But If you like to have nice click-and-don't-think or not an informatician in your lab, it is preferable to stay on Windows.
Dual OS are not very convenient instead if you could link your harddisk to both OS
There is Bio-linux with great number of tools for bioinfo work. The OS is quite large around 1.5 GB. The tools are intergrated with it so it makes work easy. Try it.
For a beginner, who want to explore the world of Bioinformatics and Computational Biology- the Linux OS will be best, provided you are comfortable enough with manual installation of various packages and libraries etc. For many beginners, it is quite a big headache how to install/uninstall a package or software in Linux. If you can't handle these processes in starting then i'll advise you to go for Ubuntu 12.04 LTS. It is a open source platform and updates automatically, the most comfortable thing in ubuntu is automatic installation of a number of essential bioinformatics softwares e.g clustalx, pymol etc. It will simply tell you the required command to install a package, run it on terminal and you'll have the software to use in next few minutes. I found it very useful and working with it since last 5 years. Try it .... have a nice experience.
I'll go for Linux in any of its flavours. I would say most bioinformatics developers use it and is usually compatible with most applications. Besides is a free OS, and it is developed by the community. I personally use Linux in Debian or Ubuntu.
"Bioinformatics" is a small word for a Huge field. We are very unlikely to agree on what it encompass.
In any case, only a Unix OS will allow the proper command line tools.
With Mac OS X the easiest to access, albeit the least complete, it might be a good place to start:
- No need to install
- Great interface
- Familiar environment
Once familiar enough with the type of Bioinformatics the person wants, he/she will be sufficiently proficient to move to a more advanced OS is necessary.
Enjoy!
I agree with Michael Baudis. OS X is one option when it comes to high graphical bio-informatics software. It looks stunning and seamless, but, i must say there are still certain problems related to the software's compatibility with MACs. Some just crash. Eg. Hex and other graphical interface software. Apart from the above negative, OS X is the best.
Thank you all for your answers.
I personally think that MAC OS X is a good option to have a "ready-to-go" powerful system based on UNIX but for someone who wants to start a career in computational biology (so, we are talking about a long term investment) it's worth to learn Linux.
I have Linux (ubuntu) and Mac OS X, and, in my opinion, Mac OSX is much better
Well this is a very broad question:
Could you define Computational Biology / Bioinformatics?
If you work on crystallography for example, Linux might be the way to go (although you can easily survive with Windows and a Linux VM in our days). Other "computational" task, which are often used in clinics (health care?) are actually soley based on Windows. MacOS is something of a hybrid.
Personally I would recommend to deepen the "user-level knowledge of computers" on what ever OS he possess, and decides later.
Except for MacOS, you an switch between Linux and Windows rather easily (at least hardware and data-wise)
Jan
The point is how big is the file that the OS is able to handle. We were working with a 32bits Windows Vistas and our file became to big to be opened in Vista (2 Gb). We switch to Mac OS and problem solved. This is a very important point.
The size of the file depends on your desktop memory size and how the algorithm running.
My experience are manage to run 12Gb of bacterial genomic data at one genomic assambly process by using LINUX OS.
Red Hat based Linux like centos, fedora, suse are all good choice considering the compatibility to computational clusters. I have had worked with many computational center, most of them are using redhat based system, I guest it is because licensing issues. Computational works could involved clusters, so for best compatibility between your computer to clusters, its better choose one of them that the programs works with little modify. I just switch from centos to fedora(google chrome works now, and it's more fancy and you can just find a usb drive and download a liveusb tool for installing)
@Diogenes Infante
It is more because you switch from 32 to 64 bits OS that you solve your problem rather than to pass from windows to Mac OS. Mac OS is cool but it is more expensive compare to linux distribution as fedora or RedHat
For years I was a Linux and Windows addicted. Finally, I switched to Mac OS X and after more than 5 years I still very happy with that and I'm pretty sure that I won't ever go back. Much simpler life here :-)
All software you need has been ported to OS X and many others are native and free.
For huge computational tasks you won't use your PC but mainframes running under some variants of Unix system. A good practice of OS X could be of some help also in this case as a starting point.
The OS you are looking for is Windows XP, all type of Bioinfo work is easily performed using this platform.
But, for more better and good results you have to use LINUS OS, and Ubuntu is best among them.
If your friend wants to focus on computational biology, they should aspire to learn more than "user-level" knowledge of computers. For many computational biology applications, you need to get into the command line, and sometimes write a little code. Anything UNIX-like is good. Mac OS X works well, but actually I think many Linux distros are EASIER to install bioinformatics software on compared to OS X. Ubuntu Linux might be the best way to go, I think. I suggest Linux over OS X mostly because Apple has been making it more and more of a hassle to get at all the "power user" features of the command line, while Linux continues to embrace the command line and develop new ways of making it easier to be a command-line user.
I've been using successfully Ubuntu for 3+ years now (most of my colleagues also). It's them most popular, very stable and extremely user friendly Linux distribution, based on Debian, regularly updated and with long support. Most of the software can be quickly and easily installed (sudo apt-get install command). And it's free!
I would suggest Linux. Very stable, quick warming up and hardly a crash. If it's your first time with this OS and you are a windows user, try the dual boot. You can get into Linux little by little keeping two OSs. I would not reccomend a virtual machine. I tried it once on windows 7 and Linux was very slow.
A good way to get a quick start is to install BioLinux (http://nebc.nerc.ac.uk/tools/bio-linux/other-bl-docs/waystorunbl) as a virtual machine under your current OS. If you like it, you can use a more permanent installation technique later on. It is based on Ubuntu:
[From their website] Bio-Linux 6.0 is a fully featured, powerful, configurable and easy to maintain bioinformatics workstation. Bio-Linux provides more than 500 bioinformatics programs on an Ubuntu Linux 10.04 base. There is a graphical menu for bioinformatics programs, as well as easy access to the Bio-Linux bioinformatics documentation system and sample data useful for testing programs. You can also install Bio-Linux packages to handle new generation sequence data types.
i DO AGREE WITH MOST OF THE COMMENTS. Linux is the best. If u wana pursue a career in computational biology, u need to learn programming. And as Lexa has suggested, prefer biolinux now without the need to configure many programs. their ppa can also be added to any ubuntu installation to install the packages with a click each.
chk this info:
http://nebc.nerc.ac.uk/tools/bio-linux/other-bl-docs/package-repository
Computational biology means: dealing with huge data sets and files. Thus, it is absolutely mandatory to (at least have an option to) use an UNIX based system and get familiar with commands like grep and cut, with 'awk' and with the '|' pipe symbol and its meaning.
My advice are
- don't be afraid of the command line
- take your time to get familiar with Ubuntu and a small set of UNIX commands
- don't panic! Nearly every problem can be solved using Google ;)
For my experience, Windows is very limited, even if some useful softwares are available only for it. But on the other side, Unix-based OS are more stable than Windows, and there are much more available free softwares that you can use. I used both OS X and Ubuntu, and they are both good. OS X of course has many other advantages on every-day use of the computer, but Ubuntu (and Linux distros in general) is free.
For a useful topic on best linux distro, I suggest this.
https://www.researchgate.net/post/Which_linux_distribution_do_you_recommend
I agree with all, use linux, and ubuntu is best!
When I'm working with my windows laptop, then I use terminal ssh to access the linux machine. And in computational biology/bioinformatics, the tools you need are all available in linux platform, free including R packages.
I prefer Linux. You can choose your distribution, but I recommend Ubuntu or Kubuntu (KDE have some great app, as kate, which you can use to select columns) or Rocks from Centos if you have a CPU cluster.
I too prefer Linux... Some of my friends are using Biolinux.....try with this
Got to be some flavour of Linux! Try to get as big a memory as possible though to avoid disk swapping as this will kill the performance of any app your running!
Linux, cause most of the bioinformatics software tools are developed only for Linux platform
I will recommend Operating Ssytem for Drug Discovery (OSDDlinux) for researchers working in the filed of computational biology/chemistry . This is new OS your input is very importan on OSDDlinux
http://osddlinux.osdd.net/
The processes and memory in windows operating system occupy most part of it. This in turn leaves comparative less power and resource, for heavy scripts to run.
It is not the fault of windows. In fact, the reason to provide you with all possible features, windows need to occupy these resources.
In Linux based systems, the occupancy is quiet low for the processes. And hence your task can be executed with an ease.
If you have an installation disk for windows I would install both linux and windows (7). This will cover the scenario where you have a bit of software that is OS specific (typically windows).
Of course, make sure you check whether the OS is compatible with your computer e.g. I tried installing ubuntu 12.04 LTS onto my lenovo ideapad laptop but it simply did't want to boot up.
Hey Gustavo,
The only thing I did in the BIOS was switching from UEFI to legacy mode (had windows8). I didn't try anything more advanced and decided to use a linux server. Though it seems like a few people have issues with getting the z580 to work properly after a quick google search.
I also deleted 8 haha!
Yes, I switched the boot order in the BIOS so that it would boot up grub first and allow me the option of choosing my OS (dual boot with 7). The issue however still remained. My friend and I then started to play around with various settings on how Ubuntu was literally booting up (we found very short term fixes by taking off the splash screen and reconfiguring a shadow file I think) but nothing worked so we gave up.
I recommend Operating Ssytem for Drug Discovery (OSDDlinux) as Dr.Gajendra Raghava mentioned..
ubuntu is the most supported distro, so i'm assuming that compiling from source or installing precompiled packages would, in most instances, would be easiest with ubuntu. ubuntu has a couple of flavors to choose from too. if you want to go lightweight, id suggest lubuntu, otherwise test them out and see which one is best for you.
other than that RHEL distros are quiet common as well. So probably Red Hat, CentOS, and Debian in that order. Red hat will be most up-to-date repos and releases, CentOS is the second to last stable version of Red Hat essentially, and Debian is a bit more hands on developer type of OS but it certainly has some cool flavors (e.g., CrunchBang)
If you really want to go hardcore, try gentoo, or archlinux, but don't expect easy solutions / support.
Probably most importantly, you want to look at the tools you will be using and see which linux distro they support, and what repositories they update and how frequently. In that respect, ubuntu will probably always win hands down.
- Badges
- Science topic
- Similar topics
- Bioinformatics
- Bioinformatics and Computational Biology