Do you know of anyone in bioinformatics doing open notebook science?
That is, do you know of anyone who puts all of their research notes/code/data/graphs/manuscripts online for the world to see, and keeps them updated every day, as they are doing their research?
I'm familiar with several examples, but none in bioinformatics.
Thanks!
like many others(?), (my blog + friendfeed + twitter) = my notebook
EDIT: four years later
Same here: https://github.com/joejimbo which is documented at http://joachimbaran.wordpress.com/
Heather - yes ONS is easier to assess when dealing with experimental sciences. Especially in chemistry we are all trained to keep a lab notebook of the experiments we run. I don't think computational disciplines have this type of training and expectations. But if a computational researcher were to do full ONS the test would be - "can people assume that if they don't see it in your notebook then you definitely haven't done it?" Whatever level of detail required to make that question true would probably classify as ONS.
In our lab when we did docking we recorded it in our notebook in a style similar to wet experiments - the objective being that someone could follow the information given there and reproduce our work. We haven't done many runs but what we did is here.
Hi Jean-Claude. Thanks for the link to the most related part of your notebook, that's helpful. Agreed: the lack of lab-notebook training and culture is part of what makes figuring out how to do ONS difficult for me. Your ONS test sets a high bar in this realm where "trying something" often takes only a minute or two... I'll think about it. Thanks!
Hi Heather, there are many flavours of ONS (see http://onsclaims.wikispaces.com/) and not all are released in real time. I keep a WordPress blog ELN for my bioinformatics research, including protocols, archiving smallish datasets, results, code, and notes. It works really well, but I've decided to release after a delay (submission of manuscript) so I'm not sure that helps you much if you want to browse- but if you've got any questions, let me know.
Yes, exactly. I've been wondering how to best release stuff. Maybe http://www.datadryad.org/ will be the place to archive stuff related to the research that publishers can't/won't take. I'm open to suggestions though! I think that permanence is something to think about. How can I make sure my ELN is still available in 10 years time? Maybe we should be hassling publishers to take responsibility for more data types? ie including the lab book records.
Heather,
I keep a real time open notebook in bioinformatics on openwetware and host my code on github. My subfield is Comparative Phylogenetic Methods, which uses phylogenies and morphological data to answer questions in macroevolution such as identifying adaptive radiations of diversity, adaptive landscapes and changes in selection regime. the day to day is a mix of theoretical methods development, implementation of these approaches into software which will eventually be distributed to empiricists and exploration of preliminary data.
While Github versioning, log, and branches provide an excellent platform for tracking, sharing and potentially collaborating on the day-to-day code development, the openwetware notebook helps me keep tack of bigger picture questions and a more accessible forum for collaborators. Wiki markup makes it easy to embed code snippets, equations, plots and links. I write daily entries reflecting the work I'm doing, meetings, outlines of manuscripts, etc.
I've implemented a uservoice site coupled to the notebook where anyone can add requests for features. So far I do not have a release version of my software, though the source code is all available. I find Mendeley public collections with webiste embedding an great way to embed the literature I'm reading into my notebook, as this is also essential part of my scientific process. I'm still figuring out what is the most effective way to do all this myself, so I'd love to see examples from other researchers in bioinformatics and hear feedback on how my approach looks from the outside.
-Carl
I'm trying to keep an open notebook of the genome sequencing project we're doing on a Pseudomonas fluoresens strain. I'm writing up notes as an online notebook which can be found here. I make all the code I use available as git repositories. All of the sequence data is available on github for anyone who might find it useful. Finally I also blog about interesting results on my research blog.
Heather - Dave makes a good point - there are examples of Partial ONS (PONS) where experiments are reported with a delay or with only selected content. Whether that works for you depends on your objectives. In these cases you cannot assume that the researcher has not done the work if you don't find it in the notebook.
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
Hi Egon. The reason is I'm asking is that I'd like to do my research in the open, and my research is tangential to bioinformatics, so I'm looking for examples to learn from.
Yup, ONS has been focused on wet chemistry to date, but I feel that ONS (or a similar concept with a different name) is really about opening and making public one's research process and results and output as they are done... so to the extent that bioinformatics is research, some variant should be possible.
Open source is part of that but not all.
Thoughts?
Interesting question... funnily, Open Notebook Science is much more oriented at wet chemistry... so, why ask about ONS in bioinformatics?