path: root/transcripts/open-science-summit-2010/andrew-hessel-pink-army.mdwn

Hi everyone, my name is Andrew Hessel. There's my twitter account there. I am
going to talk to you there called Pink Army Cooperative, it's a drug company
for everyone. It's kind of been a mission for me to build this. My background
is that I was a cellular biologist, and geneticist, worked for Big Pharma
until they kinda sorta slow down. The whole genomics bubble was really good.
But then we had to digest that, and it really takes, it takes a lot of time to
go and make drugs. Man, we're so used to thinking it's going to take 10 or 15
years and a billion dollars. After this run of this company, I had some money,
and some free time, I thought I was going to go look at these problems in a
slightly different way. This is the result.

I have a strange take on biology. I didn't want to be a biologist, I wanted to
be a computer programmer. I got into computers in the 1980s, and everything
changed so fast, it was hard to keep up with it all. New languages, new
hardware platforms, new software, I spent all my money trying to keep up. I
thought that cells were kinda like computers too, they compute with chemicals,
the operating system has been standardized for billions of years. There are
comparisons between biology and computers. Put yourself down to the level of
bacteria. You're a complex system. Microbes are a kind of simple computer,
they have an operating system of about 4.6 megabits, they can do a surprising
amount of biochemistry, and pretty robust and it's all there. Plus they
duplicate which is really neat. This is a real quick slide about analogies
between components and biology, circuits, pathways, modules, all the way up to
a full computer and ultimately aggregations of computers, like our cells,
networks of cells working together, it's all accessible through DNA. You can
program it.

Reading and writing programs in this biological space translates into DNA
sequencing and a new technology called synthetic biology- it's genetic
engineering, but instead of a lab, you use printers. This is the old way of
doing genetic engineering- it's complicated cooking, it's like doing cut and
paste with words, you don't know if the scissors are working, you can't see
the words, it's really really slow even to put together a paragraph let alone
a novel. But when you have DNA printers, you can basically, if you can type,
you can be a genetic engineer. Today there are gene printing companies like
GeneArt, Blue Herring, and a dozen others, that will print DNA from 40 cents a
base to 60 cents a base, the cost has been falling exponentially. Anyway,
anyone can become a genetic engineer today, you still need a lab to boot it up
and do anything useful. If you look far ahead, you can see that things are
going to explode. I am co-chair for biotech at Singularity University. If
everything is changing and doubling, where are we going for 5 years, 10 years,
20 years?

I really want to see genomics become like a software industry. Design, compile
which in the biological world is just printing DNA, execute and putting that
into a cell, put this code in there, do something useful with it, something
that I can test, and this is done in a lab, to produce new applications. This
is a universal formula. It's increasingly true with biology, but this cycle
for drugs has taken about 15 years, often longer. How do we accelerate it, how
do we make this wheel go faster?

Well, I started working with this group of really passionate students in
synthetic biology, called the iGEM competition at MIT. We took kids so young
that they weren't already chipped in the head as to how to go out and do
biology- most biologists take things apart, they don't think of building
biology. That's different, that's true engineering, not reverse engineering.
We took these kids, gave them tools, and gave them some little instructions.
We used an open source model, they shared their data, and it accelerated
really fast. There were about 25 countries, there was 110 teams, this is about
6 years into the whole effort, and some of the genetic engineering that they
did was simply breathtaking, and they did it 4 months over the summer, with
poultry amounts of money, we're talking from $5k to $50k and a lot of that was
just for salaries, they did kickass work and they raised the bar every year.
And I started watching this process grow, and right now they are doing it
because it's fun, and it's really fun. It makes biology and genetic
engineering fun, almost a team sport, this is like the Olympic games of
synthetic biology. When do we get serious?

You've seen DIYbio, they are hacking hardware, you've seen presentations on
drug discovery and collaboration etc., how do we go and do drug development
collaborative? How do we do DIY biotech? This is not a toy tech. My first
computer was a toy. This is not a toy tech, it's made the cover of some of the
major journals, you can be playful with it, you can do all sorts of cool
things, like bacteria smelling weird, making them turn colors, it's as
powerful as anything in biotech 15 years ago, but even more powerful. There
are dozens of companies, like Synthetic Genomics, they are just using this for
biofuels, and this company, Craig Venter started it, and it's probably valued
at over $1B already. There are other groups doing this and coming together,
it's one of the most foundational technologies in life sciences. It's kind of
like DNA sequencing and reading was 15 or 20 years ago. We're starting to see
consultancies pop up, like Gingko Bioworks. We have the tools, the tech, we
want a fee, what do you want to make? That's pretty cool. In the Bay Area,
you're going to have the first BIOFAB, an integrated design-build-test roof,
ran by some amazing academics, people who built this field, opening it up to
the whole community that might not have all of the facilities in one roof, hey
you have a good idea, let's go do it. It's a contract biology biotech company,
and once they figure out how to do it, you're going to see some commercial
groups step in and do the same thing for profit, and it's going to be
competitive, it's going to be the Kinkos of biotech.

Now, some people are going to want to replace chemical processes with
biochemical processes. Everyone is going to have a different idea. I want a
working cure to cancer. It's just where I've spent most of my life. I don't
have cancer, I'm willing to have it, but I don't. I've never been diagnosed
with cancer. All of us have corruptions in our genetic material, it's just a
part of booting up a computer, you're going to get errors on the disk, and a
certain corruption will create a cell that happens to escape all of the immune
detection and so on, and start to grow in your body. Boils, freckles and
stuff, nothing serious enough to worry about yet. But all of us are going to
get cancer if we live long enough. WE haven't made a lot of progress in this
yet. But there are a few fascinating things. It's just a corruption. All
cancers are caused by degradation of genetic material. You'll find populations
of the same type of cancer, and when it destabilizes, it can get progressive
pretty fast, and if it stays in a lump, you cut it out, no big deal. When it
gets invasive, and spreads through the network, you have to go hunt and seek
it to eliminate it.

We're learning, you know, we're getting fine resolution now. You had something
growing off the side of your neck, and the doc would say nothing you can do
about it, and we've had pathologists who can look at cancer cells; it's like
reading someone's face and saying, it's good/bad, whatever. But now we're
getting gene expression analysis, and really really high definition ways. But
what do you do with all that data? What do you do with it? I've been saying
for years that I think cancer R&D is kind of lost. It's generating all this
data, but not finding a path to using it faster and cheaper. The knowledge
cycle hasn't changed much. Cancer doesn't even really worry me, it's kind of
slow growing. In the 1900s, if you got a bacterial infection, you were dead. A
bad one, splott, you were gone. We didn't have antibiotics. We got penicillin
and other antibiotics, and it wasn't a problem any more. This is kind of like
cancer- rogue cells growing in your body and disrupting a critical system.
There's a difference though- bacteria and you are separated by 4 billion years
of distance. You can throw compounds at the bacteria and kill it, and your
body will just not really care, it doesn't effect it. It's why we don't freak
out about bacterial infections. But cancer is a different ball game. Like,
cancer cells are really complex, you've heard this, you've seen the signaling
pathways, they are complicated compared to bacteria. This is 4 billion years
of evolutionary distance. But with cancer- each cancer is unique, it's a
product of your own cells, it's one of your own cells or more that has escaped
the program, and it's starting to evolve inside of you, and it's going to kill
you if it keeps going. The evolutionary distance between the cell and your
cancer age or something is maximum your age (not 4 billion). You are treating
yourself, you're trying to kill yourself, same species, recently diverged,
every single one unique, this is probably the reason why we haven't been able
to beat cancer. We haven't had the specific enough tools, just to focus in on
just that cancer cell.

Instead, we kind of take the "nuke it from orbit" approach. We're going to
nuke every fast growing cell, why not, it's the only way to be sure. We don't
have that specific approach. So one day, you feel a lump, and a few weeks
later, you end up getting a really difficult treatment, you lose your hair,
lining of your mouth and intestine starts sloughing off, and you feel like
crap, because you're getting an optimized poison because that's the best we've
got from the last century. And then you go into survivor mode, you know it's
still in there lurking around, they are like cockroaches, it's like antibiotic
resistance in bacteria except that, chances are with highly optimized
medicine, they threw the best they had at them, so you are just happy to be
alive. I really think we can do better. I think a lot of cancer research,
well, I think it has been lost, because we didn't know about bacteria when we
made antimicrobials, we just found stuff that killed bacteria. We didn't know
about DNA, biosynthesis of cell walls, but it worked and it cured us.
Targeting fast growing cells is 1950s, let's not do that. One size fit all
drug making, it's kind of illogical with cancer, no two cancers are really the
same. The hardest part about cancer these days is that change is hard- there's
a lot of people that have been trained to think about cancer in a certain way.
I started to think, what if I got cancer tomorrow? It's not hard, I smoke. I
would want something absolutely specific to my cancer. I don't want to know
how it's going to effect you, I want it to be gentle, I want it to tickle it
to death, I want it to be effective and safe, and I want to clear it up, if I
get a bacterial infection, I want it killed today, not in a month, and I don't
want to nuke it from orbit. The clinical drug dev pipeline is tuned to make
things like blockbuster movies. There are a tremendous amount of stakeholders
for getting things through the system. There was this great paper in Nature
that called Big Pharma on this.. 60 years of innovation in Big Pharma, it's
never accelerated, this is just the best the system does. So, maybe it's time
for a paradigm change, the paper said. What does that look like?

Ultimately, everyone in drug development, in the pharma industry, are worried
about statistical analysis of population. Oh, this person, it's neutral, and
oops, that one died. You have to calculate statistics about risk benefit. But
then you think, why are you making drugs for populations anyway? It doesn't
make sense for cancer, and if it's slowing us down that much for the approval
process, how do you change that? You might be familiar with Chris Anderson and
the long tail, and how the digital tech is changing the market place. You skip
to the long tail in drug development, you get N=1, treatment it either works
or it doesn't. So I designed a drug development system, about 3 years ago now,
that started with one, isolate and analyze, we can threw a ton of tools
cheaply at this, design a drug, it's not that hard, build it, test it on that
one cancer cell, and iterate. It's cheap. Today, if we can do it in the
BIOFAB, we don't need a company, we don't need a whole laboratory, if you make
a drug, you're going to give it to an individual, just one person, and those
results are going to go back and feed into the system. What I realized is that
we have to change the business model to do it, because Big Pharma is not built
to do this. I built a cooperative, it's just community members helping each
other. It's completely open source, it's called Pink Army, we're focusing on
breast cancer. Why not? It's the most community oriented, it'll work for any
cancer, or antimicrobials. But this process is interesting. I charge $20 per
share, it's the price of the pizza. If you are not going to invest the price
of a pizza in cancer, you might want to rethink how strongly you think about
cancer. We are trying to be the Linux of cancer, we need people to invest in
themselves. The goal isn't to make one drug, it's to make a system that can
make lots of drugs.

So, my idea is that is that we only need to treat one person to make this
work. One person, one pill could really change the world in this case. One
day, you can get advanced diagnostics, download it from the web. The next step
in the evolution of the tech, is essentially everything you need to get a
therapeutic. I really hope you join us, $20, tell your friends. If you are a
cancer biologist, a young developer, or anything - I just really want to hear
from you. Thank you very much.