Database Reference
In-Depth Information
of people trying to build algorithms to tease apart those signals in the brain
to figure out what the different conversations are that are going on in a given
time and understanding how to relate these conversations back to the larger
state of the system.
Gutierrez:
Are other people tackling other biological systems of the body
with similar data and algorithm goals?
Jonas:
Certainly. A lot of my friends do computational biology work where
they're trying to understand how proteins interact and give rise to signaling
networks. We used to think that each gene was turned into a single protein—
that was the end of the story. Now we know that a gene gets turned into
mRNA that then gets more or less sliced and diced and then turned into
proteins. This kind of slicing process—called alternative splicing—is the rea-
son why it looks like we only have like 20,000 genes in the human genome, but
we have vast amounts more of all these different proteins.
There's incredible diversity in proteins. So lots of people—such as Yarden
Katz
4
at MIT—are developing algorithms to take this high-throughput data
and understand what's actually happening and what the generative stochastic
processes are. If you take the naïve computer science view, every cell is
basically a little computer, right? It has this chunk of memory, and DNA is the
compressed obfuscated buggy binary that runs inside this complicated set of
stochastic differential equations. If we're going to figure out how this all works,
then we have to start applying better computational techniques. So, yes, it's
very much the case that there are people tackling different biological systems
with similar data and algorithm goals.
I make the perhaps slightly controversial statement that I don't think humans
are going to be able to understand biology. I think our notion of what it means
to understand something is going to have to change. We're going to have to
be much more comfortable having a complicated model inside a computer,
where we only understand parts of it. In some sense, we were incredibly
lucky with physics. The fact that Maxwell's equations are four linear partial
differential equations that explain all this behavior is amazing and magical.
There's no reason to expect that these gross bags of fluids that we call our
bodies, which have evolved over 4 billion years, are going to exhibit this kind
of aggressive reductionism.
When you watch Steven Boyd
5
lecture, he keeps referring to the 19th-century
mathematics that we all know and how this 19th-century approach to science
just doesn't work. So we have to start developing algorithms and we have to
be using computational tools to redefine what understanding is. In fact, I think
