Database Reference
In-Depth Information
Gutierrez: When did you realize you were on the right track?
Jonas: The real aha! moment for me was realizing that a lot of other people
are doing computational neuroscience. Their work was just being completely
ignored by the actual people who were getting papers on the cover of Nature .
It's an interesting question of why we have this entire community of very smart
people putting out all these papers in these low-tier journals, and over here
we have the actual experimentalists, and they never seem to communicate. Is
it because everyone is insular and territorial? Or is it because what's being
produced by the computational neuroscientists isn't of use to the experimen-
talists? Why, if I'm an experimentalist, would I go and learn all these additional
techniques if it's really not going to give me any sort of new capabilities? I saw
that happening in 2004, and then I came back to it ten years later and still
seeing the same thing is very frustrating. This has to change. Someone just
has to do this.
There are more and more people waking up to this realization, and that's
part of what motivated us to get a workshop going, because I'm going to be
wicked bummed if these data sets come out and no one knows what to do
with them. I think everyone—including the funding agencies—is going to be
very frustrated. We don't want to find ourselves again in kind of the regime
that genomics found itself at the end of the '90s where it's like, “Well, okay,
we have this data. What do we do with it? Where's this clinical miracle that
everyone was promising?” Part of that, of course, is because biology ended
up being fantastically harder than we ever anticipated. But part of it was also
that we just weren't really sure of the questions that we were going to ask of
these large data sets.
Gutierrez: What are the main types of problems that people are working on
in computational neuroscience?
Jonas: There are two main problems people are working on—one part is the
wiring of the brain, and one part is the activity of the brain. The set of wiring
questions people have—especially with this connectomics data—are: What
are the circuits? How are they organized? And what are the kinds of modules
that are connected? The set of activity questions are: What are the repeat
patterns of activity? And what do they really mean? The problem with figuring
out the answers to these sets of questions is that all of the data is incredibly
noisy.
One way of understanding the data issue we face is by imagining that you are
in a stadium and you can listen to 500 people at once. Your goal is to figure
out what's going on in the game just from listening to those people. There are
certainly some things you can tell, like who's winning, who's losing, and that
sort of thing. But if you wanted to actually understand things play by play, it's
actually much, much more difficult. Translating this example back to the brain,
we can think of cells in the brain as being individual people. So there are lots
 
Search WWH ::




Custom Search