© 2017 by Eduardo Gade Gusmao

Computational biologists: moving to the driver's seat!

December 17, 2017

 

Computers were at service of many discipline since its first widespread wave in the 50s with the famous Fred Sanger insulin protein sequence determination. At this time the bioinformatics, as a discipline, didn't exist as a label (actually it did, but meant something else [2]), but was already in application. For instance, nobel laureates Fred Sanger and Max Perutz did use computers in an almost bioinformatician way in x-ray crystallography. This, back in 60s. Interestingly, "wikipedying" a little you are always surprised by some informations, such as "A pioneer in the field was Margaret Oakley Dayhoff, who has been hailed by David Lipman, director of the National Center for Biotechnology Information (NCBI), as the "mother and father of bioinformatics." [2].

 

 

 

 

From left to right: Sanger, Perutz, Dayhoff (yes, programming with punching cards... respect!) and Lipman.

 

Nevertheless, the dissemination and formal beginning of bioinformatics, as a discipline, started when DNA sequences became more widely available. It turns out that efficient computational methods were needed for the assembly of the genome (and first assembled human genome, science vs nature, etc...). What matters in this post is that at this time, bioinformaticians were service providers. Computer scientists were basically minions (yes, the yellow ones from the movie) to experimental biologists.

 

However, a couple of decades after, a.k.a now, the complexity of data grows in an unprecedented way and they do need to be correctly treated by robust computational/statistical/mathematical methods. This transformed the role of bioinformaticians, who are now group leaders, professors and authored many breakthrough publications. As an example, many argue that BLAST deserved a Nobel award as it opened the door to many of the tools we know today. As of 2009, BLAST had 26587 citations vs 14510 for PCR.

 

Of course you can be an exclusively experimental biologist and seek the help of an exclusively computational scientist (or the other way around). Although this actually happens and works, there will always be a two-sided caveat: (1) without deeper understanding of the biological protocol, will the computational method developed encompass all the data treatment necessary? (2) without deeper understanding of the computational technique, will the experimental biologist interpret the data to its maximum extent?

 

Personally, my understanding is that a good bioinformatician is not only someone who studied both disciplines (still, often with a bias towards one or another), but a diplomat; who can talk to both pure experimental biology and computational scientists and provide a channel of understanding.

 

Of course, there are still scientists who are "purists" and are unable to see that the future of research is multidisciplinar. One peculiar discussion experienced by a friend of mine can be used as a perfect example. During a talk by a computational biologist on big data and computational methods to, let's just say, "narrow down" the search space of targets of genes for a particular biological process X; one scientist argued that without the proper experimental results, all this "computer stuff" was useless. Then, a more senior (and wise) professor told the complete story on how the process X came to be known; and that without narrowing down the potential targets with the use of computational methods the process would have taken a very long time to be discovered.

 

However, the "multidisciplinary" aspect of science is not the only thing "purists" fail to see. They are somehow lost in time and fail to see that the volume of massive data being generated is growing exponentially (check my post on the number of "-seq" methodologies I was able to compile / there is many more of that, each of which generating a tremendous amount of data). As an example, the famous graph from Nature:

 

 

Finally, with all these arguments, I'll reach my main point which is: funding bodies must shift themselves to the research time we live in. Just look around. Bioinformaticians are group leaders, corresponding authors of groundbreaking research and leading research with citation records. Not only "bioinformatics" were first used by Nobel Laureates (Fred Sanger and Max Perutz), bioinformaticians are currently receiving the Nobel award (Michael Levitt [3]); for bioinformatics as we now it today.

 

Michael Levitt.

 

So here stays my appeal to "purists" scientists and funding bodies: There is no doubt that experimental biology and biotechnology is very important and paramount for the advances we have seen; but bioinformatics is more than equations and a computational program to press a button and generate pretty graphs. It is also very important and should be in equal footing with pure biology.

 

Now, with respect to the importance of "computer programming" to university biological (not only that but all others) programs... This is a subject for another post.

 

References:
1. Yanai I & Chmielnicki E (2017) Genome Biology, 18:223.
2. https://en.wikipedia.org/wiki/Bioinformatics
3. http://www.buffalo.edu/news/releases/2017/04/037.html

 

Figure Sources:
Fig.1. http://www.sru.edu/academics/majors-and-minors/biology-bioinformatics-and-computational-biology
Fig.2. https://en.wikipedia.org
Fig.3. https://www.nature.com/scitable/content/the-phenomenal-growth-of-sequence-data-in-78429
Fig.4. http://www.buffalo.edu/news/releases/2017/04/037.html

 

Share on Facebook
Share on Twitter
Share on LinkedIn
Awesome!
Please reload

Please reload

Tags
Please reload

Archive