What Is Bioinformatics?

Question

[Egon Willighagen]

The question about the oldest bioinformatics publication resulted in Casey differentiating computational biology from bioinformatics, the latter just being about (insignificant?) IT plumbing, the former about science. Depending on the outcome of this definition, Rajarshi's question is unanswerable (or subjective).

Do you agree? Is bioinformatics indeed about coupling scripts, instead of trying to make sense out of biological information, and as such Science in it contributing to what we know about nature?

[Brad Chapman]

Egon, I did not downvote, but your question as phrased is rather divisive. "insignificant IT plumbing" does not capture the spirit of Russ' blog post; he emphasizes that there are two sides to computational work in biology and both are important.

[Istvan Albert]

FWIW plumbing only seems insignificant because usually it works so well it is practically invisible. I wish bioinformatics were like plumbing ...

[Egon Willighagen]

To the downvoters: you do realize that it makes the 'oldest bioinfo paper' also subjective, right??

[Egon Willighagen]

Michael, do you plan to also tag that question as subjective?

[Brad Chapman]

Egon I did not downvote, but your question as formulated is a bit flamebait-ish. "insignificant IT plumbing" does not capture the spirit of Russ' blog post; he makes it quite clear that both are important aspects of the work.

[Egon Willighagen]

Fair. Updated my question a bit. The blog does indicate one as Science as opposed to the other... that's what I tried (and failed) to capture with 'insignificant'...

[Egon Willighagen]

I still meet scientists everyday that believe bioinformatics is Excel... all plumbing is already solved by Microsoft. Right?

Answer 1

[User 59]

Ah the perennial question on what defines computational biology from bioinformatics. For the record I do consider myself to be a bioinformatician but do not consider myself to be a computational biologist.

I disagree about the plumbing aspect however, and don't think that was inherent in anything Casey said, he just linked to a post. I also disagree with what computational biology is defined as in that context.

Things I don't do - I don't write algorithms and I don't develop large codebases or work on a single focused development project, I think this is the role of the computational biologist and bleeds somewhat into that region where computing science and software development are part of the landscape.

What I do however is survey the output of these people, use the tools and make decisions on suitability, talk to the biologists about their requirements, and apply the tools that fit best to their data and requirements. I develop stuff 'around the edges' to streamline, automate etc. Importantly, I DO analyse the data, and report the results back and try to fit those into an appropriate biological context.

To argue that one or neither of these is science is disingenuous however. I think it's more of where you sit on that sliding scale from biology to computing science to maths. I consider myself to be someone who interfaces between the biologists and the people who develop the tools. I'm not sure I want to expose the biologists to the computing scientists directly, I am a buffer for jargon and the likely culture clash.

This is, of course, all IMHO - and merely my $0.02. Very subjective indeed! Everyone feel free to disagree :)

[Egon Willighagen]

I don't know if Casey (dis)agrees with the post he linked to... from my perspective he just indicates it is not necessarily the same, and as such marks them as different... hence differentiate.

Answer 2

[David Quigley]

I think the main intellectual distinction is whether you primarily practice science or engineering. Engineers (broadly speaking) produce physical or software artifacts to solve a particuar problem. They spend more time developing methods for manipulating the world than discovering properties of the world. I personally think of myself as a cancer geneticist rather than a "bioinformatician" or "computational biologist". This reflects my own primary interest in the subject I work on, rather than the methods I'm using. Few molecular biologists think of themselves as "pipetticians" or "stainologists". I try to write as little software as I can (still, I write a lot!) because I'd rather be learning about cancer.

Most people reading this forum-- and, by extension, I think most people who do something that could reasonably be called bioinformatics-- do some engineering and some science. Although this is not a rule, people who work exclusively as scientists sometimes regard the products of engineering (replacing microarrays with RNA-seq, improving alignment algorithms, what have you) as a necessary but dull exercise that lets the scientist get on with his or her real work. As in all pursuits, there is a distinct pecking order. Some people love engineering challenges for their own sake-- I still get real pleasure out of seeing non-trivial code work properly or expressing a simple idea well in code, so I get that-- and don't need or want to do traditional science. Others engineers get caught in a bind, though, and think that for whatever reason feel they should be doing science. This is exacerbated by the pecking order, at least in academic science.

Before I worked in a lab, I was a professional software engineer for six years, so I've lived in that world as well. Back when I worked as a developer, I had a job at an internet firm where my code contributed to the firm's primary revenue source. I've also had a job where I worked in a service unit that assisted the real moneymakers, who were lawyers. You might be able to guess which job was more satisfying and rewarding, with respect to both the work itself and my feelings about my contribution to the the job environment. This strongly influenced my choice of how to develop my career in science: I wanted to do primary research in science because that's what I thought would be the most fun and rewarding. Doing both engineering and science without getting hung up on the label has increased my happiness. I personally think that it's the future for people in our field.

[Egon Willighagen]

Can you elaborate? It reads to me right now that 80% applying engineering results is more science than 80% creating engineering results... (both with 20% establishing new measurement data), but that can't be what you mean, right?

[User 59]

Great answer, very considered.

Answer 3

[Casey Bergman]

My view is that Computational Biology is a broad discipline that includes Bioinformatics (which I define as computational approaches to study molecular biological data) as a sub-discipline. Computational Biology also includes other fields, such as computational neuroscience, which do not classically fit in the field of Bioinformatics so defined.

I share Altman's view that there is a useful distinction to be made, but don't share Altman's view that Bioinformatics and Computational Biology are "either/or" activities, nor do I hold Egon's view that Altman's distinction implies that Bioinformatics is "(insignificant) IT plumbing."

Personally I consider myself a computational biologist, because I'm biologist by training and see computers and computational techniques (like molecular techniques) as a means to an end, not an end in itself.

Answer 4

[Pierre Lindenbaum]

"bioinformatics is the study of large research databases containing viral genomic sequences"

We finally have a definition of bioinformatics. Apparently it's BLASTing for viruses. pic.twitter.com/ZlX01MA0Gr

— Gytis Dudas🇱🇹 (@evogytis) February 26, 2018

;-)

[grant.hovhannisyan]

Who is a bioinformatician? It is a person who neither properly knows biology nor programming, but is very demanded.

Answer 5

[Nicojo]

It can be very frustrating to do bioinformatics when working with biologists or computer scientists. In the first case, biologists don't think you're a scientist (even when you training is in bology!)... In the second, you're not a programmer.

Now I'll go into the field of guestimation: where did the terms "bioinformatics" and "computational biology" come from? My feeling (maybe some of you older folks can confirm or infirm) is bioinformatics came from using computers to solve biological questions... But it eventually also became standard tools, which made it loose it's scientific aura. And biologists started looking down on people doing bioinformatics, just as technicians.

Annoyed with the fact, I'm guessing that some people started calling themselves "computational biologists" to mark their difference with technicians.

Then, once this spread a bit, there were two "schools": those who stuck with the original name (such as myself: biologist using informatics tools to solve biological problems) and call it bioinformatics; and those who make the distinction.

I may be totally wrong and please feel free to say so.

I guess my main conclusion will be to say that a "biologist" is someone working towards the understanding of living things... Whether that person is a technician, a PhD student or a professor. Is the technician not a biologist because (s)he is doing what the Professor says? (Mind you, I know some "technicians" who lead their own project, but that's another issue ;)

I think that person is a biologist as well. And the same applies for bioinformatics. So why go and make up another terminology?

I have suffered from biologists looking down at me because I wasn't holding a pipette. So I can understand why some people would want to make a distinction, even if I don't agree...

And that's my 2 cents ;)

UPDATE: I just found this paper by David Roos, who's lab consists of both wet- and dry-lab, in which he uses both bioinformatics and computational biology interchangeably... And I agree with what his assessment.

Answer 6

[Joe]

Bioinformatics is converting back and forth between file formats until you die.

Answer 7

[Kevin Blighe]

A presentation that I gave to school children at Sheffield Children's Hospital (Sheffield, UK), back in 2013 as part of a 'Study Day'. I was singled out in the appraisals after the event for giving an enjoyable presentation for the children.