Samuel's Tech Blog

I've had the idea that I'd like to model the embryologic process with semantics (combined with simulations, which would be wrapped in semantics), into a general query answering system for the developmental process of a given species, but having thought that too many ontologies etc. are still missing, such as for expressing time and space.

Well, in addition to realizing that there are ontologies for describing systems biology models, simulations, and simulation results, already (see blog post), I now also realize that work on the above mentioned ontologies (for time and space e.g.) are there as well, and that tons of work along the line of integrate biological knowledge and systems level simulations, is already done (though they don't seem to address embryology specifically, so far).

Gotta think of what that means. Most probably somebody will soon address the embryology process as well, using the mentioned groundwork, and so I don't have to do it! =P

I was at EBI last week for PhD interview (though I unfortunately failed the IAA test, and did not enter), but in addition to the opportunity to see EBI, I got to know some interesting stuff.

What do you think of that title? :) To me it sounds like one of the (many) natural next steps forward for Bioclipse sometime in future¹.

Explicit knowledge is too expensive

There are lots of things that can't be answered by a computer from data alone. Maybe the majority of what we humans perceive as knowledge is inferred from a combination of data (simple fact statements about reality) and rules that tell how facts can be combined together to allow making implicit knowledge (knowledge that is not persisted as facts anywhere, but has to be inferred from other facts and rules) become explicit.

One can easily imagine though, that storing every single piece of knowledge that could be stated, as an explicit fact, would require more storage than can probably ever be made available in this universe.

Simulations can make knowledge explicit, from first princples

It is not too hard to come up with some processes which are just too complex and involves too much variability² that it is unrealistic to try to capture every imaginable state of of that system or process in explicit facts. Instead we must seek the "first principles" that defines the process, and through simulations make explicit any knowledge we are looking for, at the time we need it (one can of course cache often accessed knowledge).

Agent based simulation seems highly interesting for biological and/or molecular systems, which are too complex and "high dimensional" to be successfully simulated solely by mathematical means.

Stochastic simulation use to be the way to go then, but it seems agent based computing provide an even more general, and powerful paradigm for simulating this kind of systems.

In light of this, I was delighted to find Mason, a free (how does the "academic free licence" compare against LGPL etc.?) Java based software for agent based simulations, seemingly with many characteristics that make it good for integration in other software (Of course I'm pondering bioclipse integration here)

It seems to be quite working ... see the Conway's "Game Of Life" implementation, further down on the page :)

And, in case MASON is not the right answer to every question, they provide a shortlist of other interesting agent-based simulation software.

This site is candy for anyone interested in computational molecular biology in general and developmental biology in particular:

• http://celldynamics.org

Haven't assessed the usefulness versus "eye candy" ratio though ... :) ("don't go for what the eyes see ..."). But at least they have developed a (seemingly) nice network analytic tool (they claim it is the first general purpose one, for constructing and anlyzing gene regulatory networks), Ingeneue, that is Java based and Open Sourced, so could be a nice Bioclipse plugin too sometime in the future! :)