For bioinformatics, which language should I learn first? 

I agree learning how to analyse data sets is beyond my scope at the moment but stepped up my game as it's really an important skill to have.

So glad to hear you are finding this useful!

Nice! I will add some Python For Biologists to the stack too :) Is it a class you are taking in school or are you planning to teach yourself?

It will be in school though I think I'll try to take a look at it beforehand !

I pre-learned some things in school too, especially for coding because it's just kind of fun. Codecademy is a good place to start for learning Python more generally, though I'll look into filling the bioinformatics gap with some new videos. Let me know if you come across any specific issues or questions I can help with!

It is fun, yes ! Thank you for your advice, I'll try to take a look at Codecademy (maybe in Java first since I will be doing some soon during an internship) but if I have questions, I'll ask them :)

If you don't mind me asking, what major would you recommend for an undergrad student interested in bioinformatics?

Glad to hear from you! I use Sublime Text myself, and I've heard a lot of good things about Atom recently but haven't tried it out yet. Good luck!

Thanks for the suggestion! I will see if I can add some videos on that to the stack.

BASH!!!!

That's great! You can definitely stick to R, since it can do much of what Python can do too. I just use both because there are some things I find easier to accomplish in Python, like parsing through large files without reading them into memory. It is possible to do these things in R, but sometimes it is easier to accomplish things in one language versus another. I like Python better as a programming language than R, so I only use R where I find Python falling short, like for making figures in ggplot :) What made you want to get into bioinformatics? and how is it going in self-learning mode? (That was what I did too!)

Thanks for your comment. I totally agree. If the goal is learning programming, definitely start with Python instead of R. R is extremely useful for statistics, plotting, and specific biology-related applications (bioconductor packages are great) and the data.frames may be easier for people to understand coming from Excel, but it is not structured like most languages. I use R more superficially for those applications, and Python for everything else. Python is definitely a better place to start for people who want to dig deeper and learn programming. You make a good point: I didn't say enough about whether beginners should start with R or Python if they are going to learn both, so I will add here that Python is easier to learn and will make computer science as a whole just make more sense. If R is frustrating at all, switch to Python and come back to R again later :) Bash is also a bit difficult. Bash has superpowers, but is a bit antiquated in its syntax around basics like loops, if statements, and math calculations. Python is easier to learn programming with.

I totally agree with you, and I work the same way. If I could make a time line for learning programming for biology i'd put it as follows: python scripting; file hierarchy (cd, cp, pwd, basic sed in bash); python packages(argsparse, numpy, advaced slicing of data structures); R base. Then people can either go to more biology specific packages (DESeq, ade4, bioconductor), or more into "data curation" and go to python pandas and or makesnake. Bash comes on little by little, and is amazing ! But a bit hard to grasp at first.

Awesome! It is a great time to go into bioinformatics right now! It sounds like you have a great start. When it comes to learning SQL, I would only pursue that if you have a project to apply it to immediately. I have found that I retain only what I use for my projects, so in the journey of learning bioinformatics, I highly recommend doing it through projects. This can be projects you have in school, in your own research, or some that you make up in order to learn a new skill. Where are you at in school/career right now? And what made you want to learn bioinformatics? Thanks, and good luck to you too!

Hi, I am not the guy who asked you that question but more or less I am at the same point. I am completing my MSc in Molecular Biology simultaneously with 2 years of traineeship in bioinformatics, doing an approach to genome wide studying ischemic stroke using SNPs. I am willing to apply for a bioinformatics phd program, still I have problems in learning R and python that good to develop something. I am only doing stuff related to my research. Do u have any advice ?

hey, i also have a degree on biotechnology. Plz share something coz i wanna learn bioinformatics too. email. basanta.dhungana.2053@gmail.com

If you love coding in Ruby, then by all means use it. I switch between languages to suit the needs of each project, so if you find that Ruby is missing something you can always implement it yourself or just hop back to a language that already has a package for that. I haven't tried Bioruby myself, so I don't know how good it is, but let me know when you find out!

Thank you a lot! Especially for a popularization in the light way such interesting, but a bit closed (in my feelings) area. Great job!

That is a really great question. Applying the skills you learn is the hardest part and also the most important. I think the best thing you can do is to start small and break up what you are trying to accomplish into manageable pieces. If there is a mini-version of your vision that would get you half-way there, start with that first. And you can use google to search for specific answers to your questions, or if you are completely stuck on where to start, try asking on Biostars, or if the issue is programming and not specific to biology, on Stack Overflow. I am also happy to try to point you in the right direction if you have a particular question. The idea of making a small program is a good one, that I may cover as a video in the future, but for now I can just tell you that for Python I use a package called argparse to turn my python scripts into command-line programs. You can see an example of that in use here: github.com/MariaNattestad/Assemblytics_web/blob/master/bin/Assemblytics_summary.py Thanks so much for your ideas, I will definitely try to incorporate them in future videos, and let me know if you have any specific examples of where you say you get badly stuck applying bioinformatics to practical situations.

OMGenomics Hi Maria. For some reason I never got notified about your reply! Thanks a lot for the reply and some really useful tips and pointers. Hopefully I'll start to write some very basic scripts which I can use on my day to day work. Thanks for your offer of assistance, much appreciated. Look forward to seeing new and interesting videos on this channel!

Hehe, I love web apps too ;) ,but I'm impressed you started with them and didn't give up! What was your first web app and how did you make it as a beginner?

Very nice! What languages or framework did you use? I can see some Bootstrap :)

You are right! I used Bootstrap for styling, for the rest I built from scratch with a bit of PHP, a bit of R and MySQL database, deployed via Nginx. I can see your web page is well designed! By the way, data visualization is one of my research interests, too!

Cool! I hadn't heard of Nginx before, but it looks interesting as a way to handle web apps with a bunch of small requests. For visualization, are you doing D3, WebGL, R Shiny, or something else?

Check out the videos on Ben Langmead's channel for alignment algorithms: ru-vid.comvideos

I started out with an undergrad in biology. I wrote up a bit of my story here: omgenomics.com/about Highly recommend adding CS skills to any career path, and of course especially for us biologists!

I really appreciate hearing from you! I totally know what you mean. I learned Visual Basic in high school, didn't touch programming for 6 years until my last year of college when I decided to do bioinformatics. And then the intro to CS course was in C++. I picked up bits of Matlab, R, and Python for various class projects the last semester, and it was so much easier than C++, it was mind-blowing. Python is still my favorite :) Check out the free Programming in R course from Johns Hopkins on Coursera (www.coursera.org/learn/r-programming), when you want something more advanced than my ggplot course I mentioned in this video.

Thank you for your guidance I am looking forward for more videos on your channel especially more on bash. One more thing to ask which language is better for the analysis of micro array and NGS data ?

Either is fine, depending on what you ultimately want to use the skills for. It's nice to work in the same languages as the people you need to collaborate with, so keep that in mind too.

Hi Carlos Most of the ground-breaking work in human evolution is coming out of sequencing and analyzing genomes, which involves some sophisticated bioinformatics work. A few examples: Awesome use of PCA in "Genes mirror geography within Europe": www.ncbi.nlm.nih.gov/pubmed/18758442 Recent NY Times article on Neanderthal evolution: www.nytimes.com/2017/07/04/science/neanderthals-dna-homo-sapiens-human-evolution.html?mcubz=0&_r=0 Lots of work by Adam Siepel, Svante Pääbo, Janet Kelso, and others to follow here. Bioinformatics is involved in all areas of genomics. Best of luck with your masters! Maria

My goodness, I could swear I've replied you before!! Thank you very much for your kind and informative reply!! I will take a good look at the suggested references! If you allow me a parallel question: Does bioinformatics plays in important role in System Biology? Once again, thank you very much.

I agree that Perl may still be used a lot, but I wouldn't say there is any such thing as an industry standard. Perl is not an easy language to learn first, and here I'm thinking about biologists, not CS majors. For CS majors, they already have their first language (it may be something like C++), and they are in a better position to look at a few different languages and pick something they grok easily. Biologists who are actually trying to pick the first language to learn should not start with Perl. Just looking at the syntax in Perl would make most beginners say "never mind, I can't learn this", but Python is much more approachable and therefore a better first language for biologists.

The first computer language that I learned was Perl and I learned it for a project, and it did a pretty good job parsing files which I needed to do for a publication that I'm releasing soon. Perl can do a lot of great things. There are a lot of cool parsing things that I can do in Perl that I don't know yet how to do in other programming languages, but it doesn't have any native plotting functions, and its plots don't look as good as in R, nor is web development a simple thing to do in Perl. Personally, I can't stand Python because I love brackets and semicolons, and Python punishes me for hitting the space bar one too many times which irritates me greatly. High level programming languages like Python, R, and Perl are popular because they have built in functions to accomplish any task, and you can write your own functions, too. But I think learning C++ is actually a great thing too. If you want to traverse a two dimensional array, you can easily make a matrix in R. But, you didn't really appreciate the beauty of it until you loop through your arrays. I think the more base functionality of C++ encourages people to write their own functions and not rely on builtin functions, so you can be a more creative web developer, and more power to you if you want to use the more popular programming languages. But code can get ugly when you are always importing libraries which are not very well documented and may expire in a year because you weren't conditioned to write your own functions. Part of why Perl is so hard to read is because you can solve a problem with crazy special variables and squiggles. There are ways to program in Perl more clearly as if you were programming in C, but because it gives you the ability to take complicated shortcuts, it's become a hard language to read thanks to the laziness of its programmers taking shortcuts. Still, I do like the way that it handles all the basic loops in the C and C++ styles. Sometimes I write code in Perl because of how easy it is to model the many C looping constructs. I gave C++ a shoutout because as I've seen it encourages you to write your own functions from the ground up, as opposed to staring blankly at your computer because there are no built-in functions or libraries which can accomplish your particular task.

Of course I know about Matplotlib, and you can use Seaborn to actually make those plots look better in Python. There is also ggplot in Python which lets you use the grammar of graphics, taken from ggplot in R. I am recommending R for biologists because it is easier to use than Python for data analysis when people haven't done programming before, mainly because the built-in data frame hooks up perfectly to ggplot and makes much of data analysis and plotting very easy.

Thanks a lot, for your rich answer.

Hi, i also have same confusion. If u get any thing plz share. email:- basanta.dhungana.2053@gmail.com

From my own experience I can say that they expect you to have at least some familiarity with bioinformatics tools and softwares, statistics, and programming.

@Infant Yehoshua .yes..and u are from ?

@Infant Yehoshua ..where?

+1

it helps if you first have a project to work on. you can start trying to learn R to do your stats on. If you are able to do that by the time you finish your undergrad, you've already got a good headstart. You could also start to teach yourself some Python, although, again, it's easier with a project.

I am sure bioinformatics has room for freelance programmers just like in tech, and I have done a bit of that myself through people I know from my graduate school years, but I'm sure there is a lot more demand out there, especially from academics in biology who don't have bioinformatics resources at their institutions.

It's hard to run analysis pipelines and tools like aligners, variant-callers, gene expression analysis, samtools, bedtools, etc. if you can't use the command-line. So linux/Unix is a basic skill that you will probably need in bioinformatics.

I would not recommend Perl to my worst enemy.

Absolutely! Also physics, computer engineering, math, pure biology, or pure neuroscience. With a bit of enthusiasm and a few projects under your belt, your major doesn't even matter -- unless it's far off like art history :)

Hang there mate. Our company spend more than 8 months to recruit a bioinformatic guy. He quited in like 2 months. Got a better job. ....get good at bi. There is a big demand. All the best. Get a good headhunter!!!

I am not an expert on the international labor market, but I do know that there are many PhD programs around the world that take international students. Getting in through a PhD or Post-doc (if you are already in your PhD) is likely somewhat easier than through a biotech company. If you have experience in bioinformatics and data then that definitely helps because bioinformatics is more in demand than regular biology. Check out the OMGenomics Facebook group (facebook.com/groups/396810934054265/) and see if anyone in there knows more about international job hunting :)

Maria Nattestad Thanks!

Sure! I have a list of resources here: omgenomics.com/resources/ that includes some Coursera courses in both R and Python. If you want to learn R, I suggest you follow a series I'm doing right now about Quick Wins in R. You can follow that by subscribing and getting the cheat sheet at omgenomics.com/R

ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-YfO28Ihehbk.html ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-_uQrJ0TkZlc.html ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-_uQrJ0TkZlc.html

Sweet! I'll add some bash videos to the stack :)

Hey, thanks for the suggestion! I posted the first video on bash today: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-EMaFdfIlK58.html

Just take a challenge with LFS (Linux From the Scratch) and you'll be a God in a shell. Really.

I don't think C# and Visual Basic are easy programming languages. They are at least as hard to learn, if not harder, than something like Python. I would probably recommend that you check out some python tutorials online, like on codecademy, and see if you like working with that language.

OMGenomics actually they are considered as user friendly programming languages even much easier than java. learning this language helped me to understand how current bioinformatics programms work, I'll try python to :D I used to understand new topics deeply so I was curious about programming languages.

My first language was Visual Basic, which was back in high school, so I know it's supposed to be considered easy. However, I would say that Python is the most approachable language I have learned so far, and one that allowed me to be productive in data analysis fastest.

Python is considered a language that is catered to the machine learning as well as various sciences due to its high level nature and the large number of useful libraries available. This is the reason why Python is favored over any other language in these fields. In addition it's sufficiently efficient because it leverages a plethora of C libraries.

Tons of online resources are available for learning Python, you can Google it and check out other RU-vidrs who focus on teaching Python. :)

@@OMGenomics Thank you so much, I really appreciate that

I'm excited that you're excited, but wow, that's a big question. I hear medical school is pretty difficult, so if you want to succeed you may have to focus while you're there. There is certainly a lot to be gained from combining medicine with bioinformatics, but everything around medicine is much more structured from a career perspective than the bioinformatics. If you want to combine them, you may have to focus on going through the official medicine requirements first, and then when you have the chance to pick residencies or clinical labs to work with, you can choose something related to bioinformatics. It's your whole career we are talking about here, so I think that's the best I can do :)

nice thank you :D

You may very well be able to get a job in the field without any additional education because the skills you do have are well sought-after. Some biotech companies are getting into the longevity space specifically, and most disease research and biotech work would factor into longevity as well, for instance through attempts to cure cancer. You could get a broad feel for the field through podcasts like Mendelspod and online courses, and just start applying to jobs in biotech as a software engineer. If you want to get a few projects under your belt first to experiment with bioinformatics, there is a ton of publicly available data that you can run machine learning or statistics on. The applications that C/C++ are most commonly used on in bioinformatics are genomic alignment and variant-calling, and anything else that is data or compute intensive, usually starting from sequencing data (FASTQ files). I know many companies and research labs are looking for bioinformaticians right now, so I think you're in a good position for the future.

Thank you so much. I deeply appropriate your reply. I do agree, that any research in the field of Bioinformatics is a huge part of curing aging, by taking care of smaller sub-parts of aging. I am trying to refresh my statistics and big data at the moment as both subjects were not a huge part of my previous work. I mostly worked on low-level hardware, and low-level C++ for computer graphics and physics simulations. I am also skilled in electronics and programming embedded devices. SO I am interested in combining my low-level C++, embedded and math skills and add up some Statistics and Big Data, with Neural Networks (A.I.). I will look at that Mendelspod website. Could you please advise some concrete online course, you mentioned above?

These are some of the online courses I recommend you check out. Different people love and hate the same course, so feel free to skip around and just take a peak at what you find interesting. Bioinformatics is a huge field, so you don't have to love all of its subfields, but this should give you a taste of a few different perspectives: For genomic data science, check out this specialization from Johns Hopkins: www.coursera.org/specializations/genomic-data-science Within that specialization: For some Python algorithms related to DNA, Ben Langmead's course is a good introduction from the computer science perspective. www.coursera.org/learn/dna-sequencing For machine learning, Andrew Ng's course is a classic that I highly recommend: www.coursera.org/learn/machine-learning For hard-core algorithms in bioinformatics, check out UCSD's specialization on bioinformatics: www.coursera.org/specializations/bioinformatics You can watch the lectures for all these courses for free. This is hidden where you click to enroll in a particular course, and when it pops up with 7-day free trial, at the very bottom you can click the "audit" link. Then you will be able to follow along and see all the lectures. They have seriously tried really hard to hide this lately. Luckily for you, most of these courses are made by CS people, so you should be able to speak their language :) It's a bit harder to find courses in bioinformatics created by biology-background people.

Thanks a lot! I am on it!

SirTrollingham Not at all. I’ve never seen anyone use Haskell in bioinformatics, but if anyone has examples please leave them here :)

I haven’t actually had any use for traditional databases in my work, so I haven’t needed it, but if your research project needs it, why not? :)

I have some videos on bash already, check the videos on the channel.

if u get something plz share.

SQL main purpose is axed toward databases, i'm not a bioinformatic expert but it doesn't seem like a great trade

Yes, Python and R will both do simulations nicely. I would pick Python myself, since I find it more straight-forward than R and would have to Google less to figure it out.

I can tell u that u may need a bachelor knowledge of biology to become a bioinformatician, just because u need the basis, then u will learn something specific only related to your project.

If want to do bioinformatics you should know about amino acids name for ncbi software

what was your bachelor background?

I would suggest you start by learning a bit about biology yourself, and try to understand what kinds of data analysis biologists are trying to do. For instance, you may find that showing them how to do statistical tests and plotting is very helpful, but that almost everything about algorithms will go largely unused. Data wrangling is also quite helpful, including filtering, reformatting, and dealing with missing data. If many of your students are working in genomics, check out Ben Langmead's course on Coursera on Algorithms in DNA Sequencing -- www.coursera.org/learn/dna-sequencing.

Thank you very much

If you are starting out at bachelor's level get a bachelors in molecular biology or biochemistry. As pointed out in the video start learning python,R and bash. Learning Java/C++ will give an edge in terms of developing software development skills. Also learn statistics and if you want to get into really deep, getting into machine learning will be give you an extremely useful skills in terms of data analysis and building models. I am currently getting an MSc in bioinformatics at University of Leuven in Belgium. Try for this program if you want to get an MS in bioinformatics. It's challenging but worth it.

You can, but if you just want to be a programmer or web developer it is faster and easier to go there directly, for instance by doing one of those coding boot camps. Programming is easier than bioinformatics, which at the very least includes programming, data science, and biology. I had already learned biology in college, so at the beginning of my PhD, I was focusing on filling in my programming skills. Since a lot of biologists are now in the same position I was, filling in those programming and data science skills in a way that is useful for biologists is the focus of this channel.

thanks for.the reply.actually I am not sure about jobs after bsc hons in bioinformatics so at least I can be a programmer if I cannot.find a job in the field thanks for the reply cheers

Biotech companies are definitely looking for good bioinformaticians right now, but yes you should also be able to find a programming job, (assuming you ended up learning programming in school). Good luck! :)

Heeey!! Good luck starting your RU-vid channel! Keep me updated, please :)