[2009-program-committee] 3rd person for Workshop 3 on Neuroinformatics2009

[Gaute Einevoll]

Hello Andreas and David;

I still haven't invited a 3rd person for our Workshop 3 at Neuroinformatics 2009, see http://www.neuroinformatics2009.org/workshops/Workshop_3/ .

Following Andreas' suggestions we have, however, two other very good people already on board:
Wulfram Gerstner and Arnd Roth. They should be able to present simplified and biophysically detailed
models, respectively, in an excellent way.

I believe the choice of model abstraction level is not mainly a matter of taste, but
can and should be addressed in a scientific way. Other parts of computational biology face the
same question, and I thus think it could be an idea to invite as a 3rd speaker a computational biologist outside neuroscience
who have thought about this question.

In an email to David on December 16th I wrote the following:

>One person I have thought about is James Sethna at Cornell. He had a paper in PLoS Computational Biology last year >(http://www.ploscompbiol.org/article/info:doi%2F10.1371%2Fjournal.pcbi.0030189)
>where they investigated 16 models from the systems biology literature and found that all were "sloppy", essentially meaning that >the estimates of the parameters were uncertain. On the other hand the models could nevertheless be able to give good predictions.
>I found this article quite interesting, and it seems to me that it should be relevant for the question of identification of >neuron models and thus right model complexity. Sethna is a physicist at Cornell, and from his webpage I see that he is doing many >interesting things. But I have never met him, and I don't know how he is as a speaker

A couple of days later, David wrote:

>I looked briefly at the paper involving Sethna. Has he also worked on neural models? - a presentation just criticising Systems >Biology models might look too onesided!

To comment on this: I don't think Sethna has worked on neural models. The way I think about this 3rd speaker, this is not so important. However, we should try to avoid a very opinionated and onesided person. As I read the paper by Sethna and colleagues they were not completely negative to the "systems biology" models they explored. I think the main message was that it is very difficult to find good estimates for individual parameters by fitting to experimental data. Nevertheless, one could still identify models which could predict the outcome in new situations. [Personally, I am confused about the term "systems biology". As you probably know, Erik just wrote a paper "Why are compuational neuroscience and system biology so separate?" in PLoS Comp Biology last year. I am a leader of the "Research School of Systems Biology" at our university, and I always assumed that the computational neuroscience we and others do fit well within "systems biology".]

David also mentioned that Nicolas la Novere from Cambridge. His profile can be found at  http://www.neuroscience.cam.ac.uk/directory/profile.php?lenov
In addition to doing computational neurobiology he is also involved in the development of the "Systems Biology Markup Language (SBML)". He seems to do good stuff, but I don't know to what extent he can shed light on the question of choice of model abstraction level.

Another possible person is Peter Hunter from Auckland, see
http://www.bioeng.auckland.ac.nz/People/people_display.php?people_id=353
He has for many years been developing a "multilevel" model for the heart going all the way up from ion-channels through membranes of heart cells to field-equations modelling the mechanical movement of the heart muscles. I think the question of "model abstraction level" has a lot to do with trying to bridge description at different length and time scales. So I think he could give an interesting input. He has visited our university a few times, so I know him a little bit. He is a good speaker. Hunter is also involved in the "Physiome project" (http://www.physiome.org/) and the development of CellML (http://www.cellml.org/).

We have to invite the remaining speaker quite soon. My suggestion is thus that I write to Peter Hunter and tell him about the meeting and what we are looking for in a third speaker. If he is interested and feels that he has something to contribute, I suggest that we invite him.

Best regards Gaute



%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Gaute T. Einevoll, Physics
Department of Mathematical Sciences and Technology, Norwegian University of Life Sciences
P.O.Box 5003, 1432 Aas, Norway; ph: +47-64965433, fax: +47-64965401, mobile: +47-95124536
Home: Plataaveien 9, 1369 Stabekk (+47-67580805)
e-mail:  Gaute.Einevoll at umb.no   URL: http://arken.umb.no/~gautei/