Get your grass skirt on, the proceedings from the Pacific Symposium on Biocomputing (PSB) 2006 have been recently put online. There are eight papers about semantic webs for life sciences, two of particular interest to anyone who is building and using ontologies in bioinformatics, and didn't make it to Hawaii.
Building ontologies quickly: Zero curation
The first, a paper by Ben Good et al at the University of British Columbia called Fast, Cheap and Out of Control: A Zero Curation Model for Ontology Development, evaluates a method for streamlining and automating ontology development. In the words of the author:
While ontology-building has proved to be an expensive and labour-intensive process using most existing methodologies, the rudimentary ontology constructed in this study was composed in only two days at a cost of only 3 t-shirts, 4 coffee mugs, and one chocolate moose.[!]
Thats definitely fast and cheap, it will be interesting to see if these sorts of techniques can be scaled and extended to build useful ontologies for other areas of bioinformatics.
Using ontologies to guide users
The second paper by Kevin Yip et al at Yale University, called SemBiosphere: A Semantic Web Approach to Recommending Microarray Clustering Services comes from the people behind YeastHub.
The paper outlines an experimental tool called sembiosphere: which recommends web services based on a small ontology of clustering services.
In my dreams I'd love to see something like this to help recommend services from the ever-growing number of bioinformatics Web Services. For example, lets imagine a non-expert bioinformatician has an uncharacterised DNA or protein sequence. What tools can they use to characterise the sequence? To what extent can an ontology help guide and inform the user? Can an ontology be used to reconcile the inevitable heterogeneities between services? The authors of the paper seem to agree these are interesting problems:
There are some large-scale projects aiming at facilitating complex in silico experiments that adopt semantic web technologies, including myGrid [e.g. Taverna] and BioMOBY. The current study is related to, but not covered by, these projects. It is an interesting future work to see how the ideas of the current project can be applied to their general frameworks.
I'd like to +1 that.
