Should I Do A Dual Master Degree In Both Bioinformatics And Cs?

Question

[dannytran01]

Hi,

So I will be starting my bioinformatics MS program in the fall after receiving a dual BS in Biology & Microbiology. As I reviewed my school curriculum and looked at bioinformatic job posting of all variants, I feel like my master program will not be sufficient when I go into the work field. I'm highly considering signing up for an online CS-based MS program as well to cover deeper parts where the bioinformatics program seem to be lacking: web/app development, software development/algorithms, efficient code writing(?), assembly language(?) and other stuff.

Am I just being paranoid or is this actually a good idea to do? Any advice? I eventually want to work in the industry, but if academia happen, then so be it.

Thanks a lot!

[Obi Griffith]

I think if you want to do bioinformatics (in academia or industry) then your plan is excellent. You have a good foundation in biology already. A bioinformatics masters will presumably introduce you to bioinformatics theory and applications. But, if you improve your CS skills on the side you will be a triple-threat. Although, a formal CS masters may not be necessary. I would definitely hire someone with your proposed CV.

Answer 1

[Gabriel R.]

Computer science is vertical, molecular biology is horizontal. I mean by this that CS rests on very few concepts:

1. Algorithm/data structure design and analysis
2. Complexity/CS theory/classes of problems
3. Computer architecture
4. Programming language and properly crafting code

That's pretty much it. Everything derives from that and understanding them properly is difficult and takes years. Biology has no real central core concepts, it's a hodgepodge of a plethora of concepts that are, relatively speaking, easy to understand. My advice is to do a CS-based program since acquiring new knowledge in mol. biology will come with time. Furthermore, you may learn about the genomic landscape of mice during your first position and learn about developmental biology in your second position. So learning CS concepts will serve you a lifetime but biology concepts will change from job to job.

TL,DR: Do a CS master's, you'll learn more useful concepts than a bio biased bioinformatics master's.

Good luck !

[Istvan Albert]

strangely enough, based on empirical observations I have come to believe that the exact opposite may be true. It is the programming and computer science that is easier to learn on the go, and other sciences are far more difficult to acquire on your own.

I have met a large number good programmers that have never taken a programming or comp-sci class at a degree granting institution. Yet I have never met an accomplished biologists, chemists or physicists that would have acquired that skill without extensive formal training. I am not saying there aren't but it seems to be far less common.

I think programming requires more innate talent than other sciences. It is closer to art than it is to calculus. An analogy would be that with training anybody can learn how to solve a differential equation, but only select people can learn how to draw a pleasing image moreover those that do require far less training.

[William]

I agree. CS skills are much more "liquid" than specific molecular biology skills. You can take you're CS skill's to every bioinformatics projects and even every data intensisve project outside bioinformatics. CS skills are a "career hedge" for life! Molecular biology / bioinformatics are much more time, place, project dependent. So first get a good basis in CS , then learn a area of interest / demand in molecular biology / bioinformatics.

[Ben]

+1 I think this is great advice and there is a shallow knowledge hierarchy in molecular biology. I agree with Istvan that it's easy for someone to pick up (e.g.) Python on the fly and start writing quick scripts, but unless they are motivated there's no reason to go and learn what's really going on "under the hood" or development-level techniques.

[Istvan Albert]

When you say "shallow" I suspect you mean "unknown".

And in that respect computer science seems more learn-able since it operates on man-made hardware and as such it is well defined moreover it obeys/follows human thought processes. Whereas the mechanism of biology are simply not well understood at this time nor is there any indication that these operate in a way that is suitable for human understanding.

I suspect that this is why it may be easier to learn computer science on your own. Whereas it is a lot more difficult to understand and put into a larger context the concepts of biology.

I think that those that think biology is easy may greatly underestimate it. From my own perspective any single time I tried to dig a little bit deeper into any basic biological process attempting to explain them at a level of certainty that I could explain a programming concept or construct at I was shocked at just how complicated everything actually is and just how much is unknown even of fundamental processes such as DNA binding, transcription, translation, DNA replication etc

[Ben]

I meant "shallow hierarchy" in that specific focussed projects don't require a ground-up understanding of living systems, e.g. If you're looking at the distribution of a histone mark you can read a couple of reviews, note its "readers and writers", pathways, interactions etc. while there are obviously unknowns it's not grounded on lots of interconnected previous knowledge (which can be imagined hierarchically). It's just another way of conceptualising the OP's horizontal vs. vertical explanation. Whereas to write a piece of software that does something interesting with the same data, you would ideally know about design patterns, compilers and code optimisation, complexity theory, style guides and best practices... most of which build upon a fundamental understanding of CS.

[Adrian]

I think biology, like all sciences, still does require a ground-up understanding to do properly, or at least knowledge of several levels both up and down the hierarchy from the level that you're working with directly. If you're studying histone marks, you ideally need to understand the biochemistry of both the biological processes and of the analytical methods; you need to be aware of everything else that you might come across when looking at genomic sequence; you need to grok genetics and evolution, and you need to understand how what you're studying is related to the bigger biological or medical problems you're trying to solve.

[Ben]

I actually think that's a good example of the kind of "shallowness" I'm talking about. You don't, for example, need to know about endocrine signalling, Drosophila development, quorum sensing, viral capsid symmetries... It's easy to extend that list but my point isn't that all biological knowledge stands in isolation, just that (IMO) biology on the whole is less cumulative than computer science.

Answer 2

[Adrian]

Where?

You won't learn to program in a bioinformatics MS (unless you learn on your own). You already know biology and the biology culture from your undergrad program.

If you want to do two masters programs, do CS and statistics. Or tweak your bioinformatics program to take mostly stats and CS classes.

[Gjain]

I agree with Adrian 100%... I would give more emphasis on statistics than CS / bioinformatics

[Ben]

Not sure what your masters experience was like but during mine we had intensive programming and stats lectures, followed by a three-month pure programming project. You seem to be saying that a bioinformatics masters is just more biology, but in my experience the exact opposite was true and we were barely taught any biology.

[Adrian]

No, I agree with you that one isn't likely to learn much biology in a bioinformatics program either. ;-)

Although there is certainly a lot of. programming and stats in bioinformatics type courses, they are by necessity going to be fairly broad and shallow. Good for an introduction, and for a bench biologist transitioning, but not getting into the more rigorous topics that the OP is justifiably paranoid about missing out on.

[Biomonika (Noolean)]

At which university, please?

[Ben]

Mine was at Imperial College London (link)