NewScientist of all places are talking about ontologies. Specifically An ontology for scientific experiments.
"Called EXPO, it can be used to translate scientific experiments into a format that can be interpreted by a computer."
Wow! translate experiments, that's impressive I would love to find out how the scrawny hand-writing, contained inside the standard lab notebook, dog-eared and drenched in all manner or reagents, gets translated into an ontology, that's more impressive than the ontology itself.
However on a more serious note, I agree with the concept, that something like this should exist. It would definately help in dissemination and analysis of data (providing the data is freely provided and in a standard structure of format), but this kind of process, an ontology for all of science, would have to have a major open development across all scientific disciplines, as FUGO (Functional Genomics Investigation Ontology) are doing, in order to be agreed upon and for the claim that it represents all scientific experiments.
Going by the article, and snooping around the EXPO site, a cross discipline development process doesn't appear to have taken place, or planned for the future (my apologies if this is not the case). And according to the article it has been tested on two use-cases, one on particle physics and one on evolutionary biology, this must be the only wet-lab science that exists these days. It would be interesting to see if the large scale collaboration effort of FUGO agrees with or already has a similar structure to EXPO
A quote at the end of the article says "Software to speed up this process could be a big boost ", I think he meant to say, "This would be impossible without software being available from the wet-lab bench, to data storage, to the publication process".
Comments
Concept good, execution?
I agree too that in principle, this is a fine idea. Imagine being able to avoid repetition or follow someone else's protocol to the letter.
I had to stifle a giggle at "journals could also insist that researchers submit papers in EXPO as well as written normally." As Greg mentioned, this will need graphical workflow tools. No bench scientist is going to touch this unless they can do it by filling out form fields on a web page. In turn, that requires people willing to go through the tedious process of coding frontends to things that could easily be achieved through the CLI if only users would learn how to use that instead.
Can you tell that I've been coding a lot of PHP/MySQL sites lately?
White paper outline
Wow! translate experiments, that's impressive I would love to find out how the scrawny hand-writing, contained inside the standard lab notebook, dog-eared and drenched in all manner or reagents, gets translated into an ontology, that's more impressive than the ontology itself.
We are currently in the process of designing Experiments 2.0, a new standard where meta data will be directly incorporated into lab books by the researchers. This will allow indexers to trawl through the lab and aggregate specific field information, bringing a new level of user experience to the lab. We estimate that two additional technicians per researcher will be enough to adequately cover the minimal overhead associated with this endeavour.
We note that this will provide a valuable boost of job opportunities to what is rapidly becoming a saturated field, by which we naturally mean the ever expanding set of recently completed PhDs with inadequate laboratory training. We anticipate that our elegant solution to the twin problems of laboratory record keeping and science unemployment shall have a lasting impact on the research community.
:-)
Interesting. Likely another
Interesting. Likely another case of competing bio-standards. I've had some experience with a subset of FUGO used in the Computational Portal Analysis System. While CPAS has a terrible name, is written in Java, has a clunky interface and being developed by a bunch of ex-Microsoft engineers it is still heading in the right direction.
The system is under the GPL, they release frequent updates, it is standards based and they have a component for describing proteomics experimental/analysis workflows. These descriptions can be imported to and exported from the system in the Xar file format (which is based on FUGO, but not OWL/RDF compliant).
I think what is lacking here is a good graphical workflow editor, specifically designed for scientists. Some of the commercial LIMS systems I've seen demoed have these.
An ontology for a robot scientist
seeAlso the accompanying paper An ontology for a robot scientist at ISMB 2006 in Brazil...