This sounds great. I was speaking with some of the local faculty here (at Georgia Tech) about hosting this "Project Mendel" idea, since there seemed to be alot of interest. However, I am very keen on leveraging the existing community support at Biostar, or perhaps even the infrastructure, to advance this project. I have already received a number of emails with some pretty nice questions, and have written up a short guide as to what I imagine these questions might look like. I am very eager to see version 0.1, and would be happy to help out in its development.
Whenever I am learning a new programming language I often find it helpful, after learning the basic syntax, to try to learn the libraries and such by solving problems from Project Euler http://projecteuler.net/ or many other websites which give various toy problems in computer science. There are many such websites. I am interested in finding (or making) a list of problems of this sort from which one could slowly familiarize oneself with the core biological, statistical and technical concepts in bioinformatics (I am a n00b at biology, but have a background in computational science). I want to emphasize the comparison with Project Euler. No problem should be excessively technical, and the early problems should assume only basic familiarity with biology and statistics. I have already collected a few questions. Can any one point me to an existing list? Should we start a community wiki page for the topic?
5 answers
I am planning to submit a grant that would allow us to develop Biostar as an independent application.
I am also looking for ideas for features that such a service could offer beyond Q&A. I find this idea most intriguing, it seems to be a natural extension to the Question/Answer format, good answers could be promoted to code example status.
Sounds great, please sign up for the biostar group. I will be looking for participants.
Seeing as there's lots of discussion around in the comments, but no answers to where such problems might already be found online I had a quick look. I would also like to say I think this is a great idea. Plenty of 'Beginning Bioinformatics' style books with example questions that could be leveraged.
There's a list of 'scripting style' problems here on the SAP community network, that were posted up in 2008.
There's also a host of stuff on Paolo Nuin's 'Beginning Python for Bioinformatics' that would make a good starting point for such an endeavour.
The 'Beginning Python for Bioinformatics' link was great. I am working my way through some of the the problems right now, but in Clojure rather than Python (which isn't to speak anything ill of Python; it would have been my choice before I learned of Clojure).
The links seems to be down, http://python.genedrift.org/ says the site was hacked in December, but four months seems long for recovery...
Check out Rosalind - http://www.rosalind.info
If your interests include processing chemical structures and reactions, check out the Chemistry Toolkit Rosetta Wiki
Log in to answer this question.
Very interesting concept! Want to name it Project Mandel, after the 'father of modern genetics'? :)
Eric, you mean Mendel :-)
I would definitely favour such a project as well. Am learning a new language at the moment and did some of the Euler problems, but they bear little to no use for my day-to-day work.
Mmm... of course :)
Agreed, maybe also something where solutions could be posted in different languages like rosettacode.org
Ahah! Of course :)
Is the Project Euler site using an easy to reuse template? Otherwise, what could be a good (maybe open) template for such a site?
Pierre: of course, my mistake... Is the Project Euler site using an easy to reuse template? Otherwise, what could be a good (maybe open) template for such a site?
I would find this site interesting if no external library was used. For "How do you read a Fasta file using PERL?" : i don't want to see "just use BioPerl", show me the CODE !
Great idea Gabriel! I like the idea of calling it project Mendel.
I'm sure there only needs to be a leader for this project to gain interest. I do think that a wiki page for would be a very good starting point to collect both ideas and problems, which the project leader could use to build the site. I would definitely love to try myself against the problems of Project Mendel lol. It would give us the occasion to try that new language we are learning on something else than algorithms for finding prime numbers :)
A more worth name would be 'project Dayhoff', in the name of one of the first bioinformaticians. Mendel was not a bioinformatician :-)
What are you asking here exactly? Do you want us to point you to projects similar to Euler for bioinformaticians, or you want to create one yourself?
I feel the bioinformatics students' community will be elated to have such a thing :-)
I'm with JRS - I think if we come up with a list of bioinformatics problems of increasing difficulty, showing analogous solutions in as many languages as viable would be a great resource