What Do You Consider The Most Trivial And The Most Challenging Tasks In Your Particular Field Of Work?
9
8
Entering edit mode
11.6 years ago

To clarify, I am a biochemist, trained in molecular evolution since undergrad and I changed my particular field of work to large scale phylogenetic analysis in the grad school.

So, to me, building and interpreting a phylogenetic tree from a family of genes can be considered a trivial task in my subfield (actually, this should be be trivial to any people with a background in biology), but to integrate functional data and increase the scale of the analysis to whole genomes and dozen/hundreds of species would be a little more challenging.

I am really curious since bioinformatics and computational biology has had a fast growing in the last years, many paradigms have changed in what is trivial and what is challenging in his many subfields.

career subjective • 5.4k views
ADD COMMENT
0
Entering edit mode

I can't agree with you when you say that building and interpreting a phylogenetic tree should be trivial to any person with a background in biology. Tell me, are you any good at Florescence activated cell sorting? It's a pretty "basic" technique in biology... I don't know anybody good at FACS who can do phylogeny and vis-versa.

ADD REPLY
0
Entering edit mode

I say that because I firmly believe that this should be a skill that any biologist can have. Tools like www.Phylogeny.fr and Mega megasoftware.net) greatly reduce the time and work need to build a phylogenetic tree. Since evolution is a core concept in biology, the foundation to interpret a tree is already explored in most undergraduate courses I know.

ADD REPLY
1
Entering edit mode

In true, for quickly build a tree, it is only a matter to run a BLAST search and from the results page click in the "Distance tree of results" link. Sure, this may not be a good tree built with the most sophisticated methods, but is a good start to a more complete analysis. I am not saying that phylogenetics as a whole is trivial, just the specific task to build a simple tree and interpreting the branching pattern resulting from it.

ADD REPLY
0
Entering edit mode

Without questioning the merit of FACS as an important technique (I know it is), we must agree with this famous Dobzhansky quote "Nothing in Biology Makes Sense Except in the Light of Evolution".

ADD REPLY
0
Entering edit mode

I play around with phylogenetics. It's not simple. And yes, I understand the importance it has in biology. But, sorry, I still can't agree with you. You're view and experience seem limited if you think that every University teaching biology goes into the concept of phylogenetics. In particular there are a lot that focus on medical issues and they really don't care about it.

Trust me, the really don't. And they do very good biology too!

ADD REPLY
0
Entering edit mode

I think we are arguing about a different point of view. Maybe due to a particular bias. I am in Brasil and here all biology courses consider phylogenetics pretty seriously, undergraduates seeking to work in medical issues generally goes to medical school. Anyway, I think that even biomedical students can greatly benefit if they start to think about medical problems from an evolutionary perspective. This will be particularly important with the coming availability of a vast number of sequenced human genomes.

ADD REPLY
0
Entering edit mode

Dude, I'm not saying you are wrong in stating that it is important (you are totally right!!!). I'm saying that you're wrong if you think that every biologist in the world understands or cares about it.

ADD REPLY
1
Entering edit mode

Are you sure you can call someone a biologist if they don't understand a simple phylogenetic tree?

ADD REPLY
0
Entering edit mode

Yes. (This is all I have to say, but I have to add more characters ;)

ADD REPLY
1
Entering edit mode

Ok, so we have different views about what a biologist is, and that goes a long way. :--) Thanks for the discussion.

ADD REPLY
0
Entering edit mode

LOL, sorry I didn't want to add anything but I have to add: please, go tell all those people working on type III secretion systems or biofilms in microbiology and all those working on p53 or c-Myc and other cancer related topics! Trust me, they don't care one bit about phylogenetics ;)

ADD REPLY
0
Entering edit mode
ADD REPLY
0
Entering edit mode

or...if you have access to Scopus: http://bit.ly/coSBWK

ADD REPLY
0
Entering edit mode

...and broad results from Scirus: http://bit.ly/cB6eNu

ADD REPLY
0
Entering edit mode

Look at the stats: pubmed "p53 cancer" and compare to "p53 cancer phylogeny" :) And YES, again, you do have some that do phylogeny! But seriously, are you calling the tens of thousands of other papers that don't care about it "Not Biological"?

ADD REPLY
0
Entering edit mode

No, that is not the point.

ADD REPLY
1
Entering edit mode

Sorry to interrupt this "discussion", but the statement that phylogenetics is easy is unjustified. But because it might be me who just cannot grasp the "trivial trees" could you Marcos explain me how I can relate/interpret posterior branch probabilities with bootstrapping values and what to do if they differ significantly? And which of the two approaches for consensus/super-tree reconstruction using majority-rule voting is more relevant evolutionarily i.e. closer to the real evolution.

ADD REPLY
1
Entering edit mode

Marcin, please consider my other comments. I didn't want to say that phylogenetics is easy. But, build a simple phylogenetic tree and interpret it can be. Please, point me where I told that phylogenetics is easy. Remember that not all problems are ultra-hard and need the most sophisticated inference methods or supertrees to be solved. I definitely wasn't trying to be pedant, pretty much the inverse. See my example about buiding a tree from BLAST results, what can be more simple? Yes, it may not be the most accurate, etc, but sometimes you only need that. That is what I am talking about.

ADD REPLY
1
Entering edit mode

Marcos +1: not everyone can understand the subtleties of a phylogenic tree. But any biologist (high school student ?) should understand ( I didn't say 'be interested') what is the underlying idea described in those trees.

ADD REPLY
0
Entering edit mode

Pierre and Marcos, I'll just emphasize the key word in your statement: SHOULD. And that is my point... You all seem to evolve in the world of bioinformatics and you understand phylogenetics. But, when it comes to their own work, many biologists (yes, I insist, they are biologists, even if they don't understand phylogenetics) in wet labs do not understand the meaning of phylogenetic relationships. Yes they understand that you can make a "species tree", but no, they don't relate it to what they do, i.e. they don't understand it and they do not value it.

ADD REPLY
0
Entering edit mode

And Pierre, I have been to high school in the french system. Please tell me when phylogenetics is covered?

ADD REPLY
1
Entering edit mode

I am sorry Nicojo, but what you are asking me for is that I admit that is ok for a biologist not understand a phylogenetic tree. This is like saying it is ok for a chemist not understand the acid/base concepts, as long as he works with nuclear chemistry. I can't agree.

ADD REPLY
1
Entering edit mode

Marcos, would we have Rock music if all musicians had to go to the conservatory and learn Brahms first? I don't think so, but I'm probably wrong... I do agree that this is a very important and fundamental aspect of biology. I'm sorry I made all this fuss over this topic. I'm usually fighting with wet lab biologists to try and make them understand that bioinformaticians are also biologists! In any case, even if I disagree, still I respect your take on the matter, and I'll even use it next time I argue with the others ;). Thanks for your patience with my arguments!

ADD REPLY
0
Entering edit mode

No problem Nicojo, it was a good discussion. All the best.

ADD REPLY
0
Entering edit mode

I'll quote "building and interpreting a phylogenetic tree from a family of genes can be considered a trivial task in my subfield". If your statement is true and indeed building a tree from a family of genes is a trivial task to anyone, then my PhD is a terrible mistake is a terrible mistake because I try to give answers to trivial questions like "what is a family of genes" or "how to interpret a phylogenetic tree".

ADD REPLY
1
Entering edit mode

Nicojo, I was introduced to genetics/phylogeny in my final year of Secondary School (~17 y.o.)

ADD REPLY
0
Entering edit mode

Pierre, je viens de chercher le programme officiel et effectivement il semblerait que c'est maintenant partie intégrante de l'enseignement de terminal S. Mais de votre réponse, j'en déduis que vous êtes bien plus jeune que moi. Cela fait il de moi un "non-biologist"? Je trouve que vous avez une vision très réductionniste si vous arrivez à de telles conclusions.

ADD REPLY
0
Entering edit mode

Are you sure you can call someone a biologist if they don't understand a simple phylogenetic tree? (I am not telling everyone should care about, I would like that, but it is unreal.)

ADD REPLY
12
Entering edit mode
11.6 years ago

Off the top of my head:

Challenging and frustrating:

  • Sifting through the avalanche of new papers and tools to find ones that are immediately relevant

Challenging and interesting:

  • Given whole-genome assays, deciding which of the hundreds of mutations are important and potentially causative
  • Integrating results from assays that provide different insights and potentially conflicting information.
  • writing code using a parallel (or map/reduce) paradigm, so that it won't take years to run

Trivial, but tedious:

  • constantly munging data files from one poorly defined format to another
ADD COMMENT
2
Entering edit mode

"Trivial, but tedious: constantly munging data files from one poorly defined format to another"

Can't agree with you more!

ADD REPLY
8
Entering edit mode
11.6 years ago

From a support perspective, the biggest challenge of my day-to-day work (rather than grand scientific challenges) remains convincing wet-lab biologists of the inappropriateness of Excel and Word as data exchange formats.

Scientifically speaking, data integration is probably the biggest technical challenge. Making disparate experiments comparable, and databases compatible would make life significantly easier.

ADD COMMENT
7
Entering edit mode
11.6 years ago
Neilfws 49k

I think the most trivial is also the most challenging: fetching and pre-processing the data in the first place, before you even do any analysis with it. I'd say 80% of my time is spent searching, fetching, parsing, munging and storing data, 20% is analysis and conclusions.

ADD COMMENT
5
Entering edit mode
11.6 years ago

Challenging:

  • Using other's code
  • understanding why a technology/paper/language/db is worth looking/testing
  • understanding theoretical papers
  • making my code reusable
  • being my own learning mentor
  • working with 'big-data'
  • (...)

Trivial:

  • working with 'big-data' :-)
  • Using SQL/C/C++/Java/etc... etc...
  • using some common resources related to by field (dbsnp...)
  • Parsing data
  • sharing
  • (...)
ADD COMMENT
1
Entering edit mode

+1; among other things I definitely agree on "understanding theoretical papers"

ADD REPLY
5
Entering edit mode
11.6 years ago

These are the big challenge(s) for me:

  1. Ask the relevant questions and find/develop the proper/satisfactory solutions to obtain meaningful results.
  2. Determine what is relevant (what really matters).
  3. Understand what and why is proper, merely sufficient or unsatisfactory.
  4. Interpret the results without bias and preconceptions.
ADD COMMENT
5
Entering edit mode
11.6 years ago
David Nusinow ▴ 260

Trivial

  • Creating lists of genes that might be of interest in a given system
  • Integrating several experiments with one kind of data
  • Talking with people in my own sub-disciplines of biology
  • Thinking of cool experiments to do

Challenging

  • Determining which genes in a list are important for a process and integrating those genes in to current domain-specific understanding in a meaningful way
  • Integrating several experiments across totally different kinds of data
  • Talking with people in other sub-disciplines of biology and actually understanding each other well enough to collaborate effectively
  • Thinking of the most important and useful questions to ask
  • Getting funding
ADD COMMENT
2
Entering edit mode
11.6 years ago

Most challenging: processing contradictory information and identifying the credible evidence.

ADD COMMENT
2
Entering edit mode
11.6 years ago

My main work deals with simulations and modelling in population genetics (for several different purposes). Data are essential to validade both.

Trivial:

  • Gathering HUGE amounts of data (I do love NGS!);
  • Parsing/converting these data to a usable condition;
  • Extracting the population genetics parameters estimates from it;

Challenging:

  • Knowing when to stop gathering HUGE amounts of data (I still love NGS!);
  • Curating data <-- The real challenge;

For a population geneticist these are truly happy times. Finally we have (possibly) enough data to validate some fundamental models and theories. But, as a great philosopher said - "With great power comes great responsibility". So, carefully curated data is the great challenge from now on, IMHO.

That's all, dudes.

ADD COMMENT
0
Entering edit mode
7.0 years ago
klemen ▴ 170

Since the last post in this discussion is more than 4 years old, I would like to hear, have any of mentioned challenges changed? Are there any other challenge you see while analysing data?

ADD COMMENT

Login before adding your answer.

Traffic: 1827 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6