The Top Bioinformatics Challenges according to Chris Burge at MIT and his colleagues are as follows...
- Precise, predictive model of transcription initiation and termination: ability to predict where and when transcription will occur in a genome
- Precise, predictive model of RNA splicing/alternative splicing: ability to predict the splicing pattern of any primary transcript
- Precise, quantitative models of signal transduction pathways:ability to predict cellular response to external stimuli
- Determining effective protein-DNA, protein-RNA and protein-protein recognition codes
- Accurate ab initio structure prediction
- Rational design of small molecule inhibitors of proteins
- Mechanistic understanding of protein evolution: understanding exactly how new protein functions evolve
- Mechanistic understanding of speciation: molecular details of how speciation occurs
- Continued development of effective gene ontologies-systematic ways to describe the functions of any gene or protein
- (Infrastructure and education challenge)
- Education: development of appropriate bioinformatics curricula for secondary, undergraduate and graduate education
See Chris Burge Bioinformaticists Will Be Busy Bees Genome Technology, No. 17, January, 2002. Available (by free subscription)
Comments
"Top" says who?
A "top 10" (11) list in a specialist publication that much of the world can't access. You have to question its relevance.
To me, a lot of these seems rather broad and vague. I'd also argue that some of them are not bioinformatics challenges, in the sense that we don't understand the computation. They are biological challenges - it's the underlying biological science that we don't yet fully understand. You can't model a process if you don't have a good idea of how it works.
This illustrates a recurring problem. Bioinformatics is no use if it isn't helping you to understand fundamental biological processes and components. This is why, IMHO, we need more people who are biologists first and programmers second, rather than vice versa. When I read phrases like "protein-protein recognition code" and "molecular details of speciation", I suspect that I'm dealing with the second group, not the first.
If Bioinformatics is what
If Bioinformatics is what http://en.wikipedia.org/wiki/Bioinformatics says, then list covers most of the challenges in the field... but as Pedro said, and I also agree, Bioinformatics is much more then Genome/Proteome Informatics. Informatics has role in all the aspects of Biology and thus Bioinformatics has to cover all that. Covering MEME [ http://en.wikipedia.org/wiki/Meme ] sounds interesting too :)
To much focus on the cellular level
I am working on 4 and a bit on 8.
The only thing I would complain about in this list is the total focus on cellular events. I am sure there are many interesting challenges in developmental/tissue or ecosystem levels for computational biology. Understanding the brain is probably one of the most interesting challenges that we have ahead of us and bioinformatics will surely play a big role.
Some minor details ... why limit the quantitative network analysis to signal transduction ? Metabolism is also very important. In 7 I would add the ability to rationally build a protein for a specific function.
Defining Bioinformatics
And so the discussion comes down to defining the term Bioinformatics (which is, by the way, not at all the same as computational biology). To me (and taking the origin of the discipline into account), Bioinformatics IS focused on the cellular level by definition. It's all about molecular biology. If you want to talk about modeling of organs or entire ecosystems, about management of patient records and image analysis: this is not bioinformatics (but it might be computational biology ;-)).
That's not a free
That's not a free subscription, if you are from outside of Canada or USA.
Thanks anyway :( I'm in point 2. and a little in 10.
Someone should tidy up this
Someone should tidy up this post, or resubmit it. It is worthy of front page attention.
Thank you Duncan.
Thank you Duncan.
So do people agree with them ? Are they too broad ? and if not how much progress have we made in each ?
Computational challenges
The computational challenges are curiously missing from this list and we still don't know how to integrate all this bioinformatics data we have. Speaking of which, Data Integration in the Life Sciences (DILS 2007) has just opened its call for papers...