Wednesday, January 04, 2006
New
discussion about signal pathways
and complex ontology
This is part of a discussion that will be moved to a Wiki page soon.
Gary
QSAR does stand for Quantitative Structure Activity Relationship but I understand that there are many ways of approaching the issue of double articulation in substructure the function analysis. Right? (smiles)
I see structure-function expression as “degenerate” in the way discussed by G Edelman [1]. We see this phenomenon also in linguistic “double articulation”.
Open logic is “my term” and means merely a logic that is "open" to the formation of the logical atoms as well as the inference rules. The soviets work on this notion and this work is still the only “open logical system”.
QAT is explained at
http://www.bcngroup.org/area3/pprueitt/kmbook/Chapter6.htm
and is infinitely simpler to understand that predicate logic, and theory proving.
I do hope we can see a way to communication. It seems that the group is rich in understanding.
You said:
Also, you should also know that there are two communities coming
together to work the same biological problem - representing pathways. It has
recently become clear that each has very different expectations of the
outcome. The bioinformatics people who
run databases want to create some basic data structures to represent mostly
existing data for use in data exchange.
The people with a stronger computer science/AI (computational
logic/reasoning) background want to solve a more general problem, which is
harder and requires more research. To
deal with this, we have recently split into 2 tracks (DX for data exchange and
SW for semantic web). Each will follow their own expectations, but share basic
requirements. Eventually, we'd all like
to have a much better representation for biological pathways than the SW track
promises.
I'm more interested in the DX track right now, since I have data sitting on my desk that I want to analyze now. When DX is done, I'll be more interested in the SW track. Alan Ruttenberg is the main proponent of the SW track and would probably be interested in hearing from you.
My discussions with Berners-Lee and Hendler has been critical of the Semantic Web approach for several “principled” reasons.
I do not wish to get into that discussion, however. I have been on the record for some time that the SW approach to “meaning” is deeply and profoundly flawed. Meaning is situational, first and foremost, and this situational aspect cannot be ignored simply because the SW formalism do not acknowledge the issue.
What I am interested in is in the DX Data Exchange track, because it is here that I see the development of data ontology which then is used to create complex ontology (as required by the EU’s new funding program – which I wish to apply to).
The DX track can USE the Protégé editor and the various forms of “machine entailment engines” which are called “reasoning engines”, but use them as a more flexible and responsive database schema generator. This is where we put the data and this is where the leading edge in bioinformatics is now, and in the near future.
Once this is done, then new possibilities arise related to “anticipation” and thus to interface design based on anticipatory mechanisms. – I hope.