Reading the article Gene Wiki conceptual overview posted at TSL on Nodalpoint some time ago, and looking in the proposal made there, I wonder if this is not already being done. There is the BOWiki Project. BOWiki extends a Semantic MediaWiki by additional functionality to allow the annotation of gene data and is supposed to allow the curation of biomedical ontologies as well. It makes heavy use of biomedical ontologies in the description of genes and gene functions, in order to allow a search of the wiki using the categories in the ontologies.
There is a paper describing the ideas behind BOWiki called A Proposal for a Gene Functions Wiki, referencing a correspondence in Nature, which was also discussed on Nodalpoint. And there is a prototype of BOWiki, which still lacks a lot of functionality, but already has some nice feature.
Maybe it would be best to pool resources and get all the people interested in such a Wiki together on one table (or mailinglist), and really get the project going. Would someone want to create a mailinglist for this?
Comments
gene wiki in ucsc
only a wiki that will be used by biologists will be successful in the end, in my opinion. I smell overengineering for all things set up for research purposes in the small corner of the bioinformatics community, whatever nice functionality they may have offer. I've recently discovered on the ucsc wiki that the UCSC group plans to add a "wiki track" to its highly popular genome browser. To me, this looks like a cool idea, as most people are working with these browsers all the time anyways and they sooner or later stumble upon the wiki annotations, comment on them, get angry, whatever, but it's a simple solution that people eventually might start to use one day...
Are you behind the BIOWIKI
Are you behind the BIOWIKI project ? Googling your username leads me to this page, and your name is listed on the project members page ?
Another point. The BOWIKI
Another point. The BOWIKI concentrates on the GO, there is no instance data. Maybe a better approach might be to put the entire GOA on the same wiki system. It might help people think about relations if they are grounded in instances of biological entities that they already know.
The problem is that the
The problem is that the people in charge of the Gene Ontology are rather resistant to collaborative ontology maintenance. I floated the possibility of collaboratively editing ontologies at last years ISMB meeting. The general response was that anarchy would prevail, not order.
There were many comments along the lines of: ontology development should only be done by ontology engineers, regular people should not be allowed near an ontology. To some extent this is true, the implementation platform often constrains the way the ontology is ultimately constructed. Knowledge of platform specifics (i.e. the differences between the various flavors of OWL) in this case is necessary.
People are in general resistant to change, scientists are even more so (yes, very ironic). Collaborative editing of ontologies is an idea that will eventually become an accepted norm in the biological sciences. I'm just not sure that time is now.
As for a combining of efforts, I agree to some extent. For it to work someone would have to take it up at an official level (i.e. contact the GO people). I would be surprised if the initiative 'emerged' from unaffiliated groups on the Internet (credibility issues).
I would also be interested to read what others think re: when and if it will happen. Interestingly Wikipedia now includes a side bar for entries that reference biological entities, check out the page for Pyruvate kinase. Sooner rather than later ?
Control Freaks
I'm also surrounded by control freaks, but none of these people would mind if one or more Wikis would allow people to attach data and comments to the "authoritative" version of the data. Such feedback would be very much welcome, and much of it would even be integrated back into the official version of the data. The tricky bit with this kind of setup would be to get the synchronization right: Don't want to fork the data, but can't spend a lot of time cleaning up the Wiki to take changes into account, either.
Not only concepts
Hi,
Yes, I work with the group developing the BOWiki. My main interest there is to use a semantic wiki for collaborative ontology curation, so I am a bit biased towards its application in this field. But the idea behind the BOWiki goes beyond editing ontologies.
The main application is indeed supposed to be the annotation of genes or related biological data (yes, this data is still missing in the wiki). The idea is that people should be able to categorize genes, search or browse through them by other means than their name. One possibility to do this is to use some of the biomedical ontologies and use their annotation or let people annotate to the concepts in these ontologies. GO is one of these ontologies, and there are others. And in the BOWiki prototype, these ontologies are only present for annotation, and are not supposed to be edited. If the need arises to modify a concept in some ontology, or add a new one, then this should be possible in a wiki (you would not want to wait until changes are made official before you edit a wiki page). But then you create a new concept, which structures the content of the wiki, and do not change the GO. As you say, this concept could at some point be taken into GO, but this is not mandatory.
BOWiki is more or less a research prototype. But its goal appears to be close to the goal of other gene wiki projects. If a collaboratively curated knowledge base of gene functions is to become reality, I believe it best to have one united effort which reflects the views and needs of a wide variety of users. This is probably not BOWiki, but some of its ideas may be of use in other approaches for gene wikis, or maybe in one collaborative effort to create a gene (function) wiki. This is the reason I wrote the blog entry here.