WIRED Videos
Science
Events

WIRED25: 23andMe's Anne Wojcicki & Stanford's Stephen Quake on Big Data and Health Care

23andMe Cofounder Anne Wojcicki and Stanford Professor of Bioengineering and Applied Physics Stephen Quake spoke with WIRED’s Cofounder Jane Metcalfe as part of WIRED25, WIRED’s 25th anniversary celebration in San Francisco.

Released on 10/16/2018

Transcript

00:00
(upbeat jazz music)
00:04
Now we transition, and we talk a little bit more
00:06
about our bodies, getting out of
00:10
social media or printing or anything else.
00:13
So I have two extraordinary thinkers and innovators here,
00:17
Anne Wojcicki who is the co-founder and CEO of 23andMe.
00:23
You did biology at Yale, molecular biology
00:26
for the NIH afterwards and UCSD,
00:30
and then started 23andMe 11 year ago?
00:34
[Anne] 12.
00:35
12 years ago now.
00:36
Great, okay.
00:37
And we have Stephen Quake, who got his undergraduate degree
00:42
in physics and mathematics from Stanford,
00:45
and then a doctorate in theoretical physics
00:46
from the University of Oxford.
00:50
You've been elected to just about
00:51
everything that I can tell, right.
00:53
National Academy of Sciences,
00:54
National Academy of Engineering,
00:56
National Academy of Medicine,
00:58
National Academy of Inventors.
01:01
Founded the Bioengineering Department at Stanford,
01:05
and now you're a co-president of the BioHub,
01:08
the Chan Zuckerberg BioHub.
01:11
And you run Quake Labs as well.
01:14
So they've both had an extraordinary career so far
01:18
but they've also both had a rather extraordinary year.
01:20
So I'm going to just start with you, Anne.
01:23
You got approval for several tests
01:26
this year after famously having
01:28
approvals withdrawn, so how good does that feel?
01:32
It's good, it's an evolution, you know I think one thing
01:35
that working with anyone, anything that's regulated
01:38
it takes some time to come back,
01:39
so we had a warning letter in 2013
01:41
so we had to stop returning information
01:43
and we've been slowly proving out that consumers
01:47
actually can get this information directly on their own.
01:49
So the approval that you mention is the BRCA,
01:52
so for breast cancer people who have
01:54
one of the three common mutations
01:56
found in Ashkenazis that account for
01:58
about 85 percent of breast cancer,
02:00
BRCA-associated breast cancer in Ashkenazi populations.
02:02
So we got that approval which was significant for us
02:05
and then obviously we have the GSK announcement.
02:09
So we announced three years ago that we have
02:11
a pretty significant, we hired Richard Scheller,
02:14
and we're putting significant resources
02:16
into drug discovery and that ties in with the mission
02:20
of 23andMe was always about helping customers access,
02:23
understand and benefit from the human genome.
02:26
And so for me one of the clearest ways
02:28
that you could potentially benefit
02:29
from the human genome is in drug discovery.
02:32
So if we can actually use the collective information
02:34
that our customers have consented for
02:36
and say we're gonna translate those insights
02:38
into some way that's actually going to be meaningful
02:40
for you, as in a cure or a treatment
02:42
for a disease, I think that's like
02:44
one of the clearest ways we can show a benefit.
02:46
So one of things that, before we even get into
02:48
the drug development that's so interesting,
02:50
is just the whole idea of why I should care about my genome.
02:53
And how much actionable information
02:55
is there for me in that.
02:57
I think you know one thing that has been fun
03:01
over the last 12 years is right,
03:02
for the people who started the company,
03:05
it was really clear to us we saw immediately like,
03:07
of course everyone wants their genome,
03:08
like it's the digital representation of you.
03:10
It's like looking in the mirror.
03:12
Like why would you not want this.
03:13
And we had a big hype, we had the New York Times,
03:17
we're on the cover of the style section.
03:19
We had spit parties, we had all kinds of flair.
03:21
And then we saw sales drop down to 20 to 25 kits a day.
03:28
And we were like, wow, we really,
03:30
we're going to have to prove out.
03:31
Like, what are you gonna do with this information.
03:34
And it's one of the things that I think
03:37
that's a passion for both of us is scientific literacy,
03:40
is that helping customers know more and more,
03:42
like why is the science relevant for you.
03:44
What does this actually mean, what is real science.
03:48
The weight loss industry, for instance,
03:50
is a huge industry, and there's all kinds of information
03:53
about what you could or could not do and 23andMe
03:56
is really passionate about making sure
03:57
that people understand what your genetic information means
04:01
for you and what really is a real association
04:04
that you should do something about.
04:05
So something like cystic fibrosis.
04:07
People who find out they're carriers for cystic fibrosis,
04:10
you want to know that information when you are deciding
04:14
whether or not you want to have children.
04:15
I met someone last night, for instance,
04:16
who did 23andMe back in 2012, 2013
04:20
she found out that she is a carrier for sickle cell.
04:23
So it was a really important discovery for her.
04:25
We find people all the time, for instance,
04:27
you find that they're carriers for the BRCA mutation
04:30
and it's life changing in terms of
04:32
how they are approaching their breast cancer risk.
04:34
Whether they're getting a mastectomy
04:35
or they're doing active follow-up.
04:37
So there's a body of information
04:40
that is truly actionable, whether something you can do
04:43
and then there's a spectacular world of discovery.
04:46
Like to me it's the beauty of the genetic code.
04:47
Like we don't know what most of it means,
04:50
and so that's what we're still trying to figure out
04:52
and understand that we have this really basic code
04:55
that represents all of life and you are human
04:58
and a different combination and you could be a banana.
05:01
And so I find that fascinating because there's not
05:05
a big difference between you and the banana.
05:07
So what is the percentage difference
05:08
between humans and bananas?
05:09
I think humans and bananas,
05:11
those are one of the things I'm not positive that
05:12
there's a lot of NIH funding looking at that,
05:15
but it's estimated about 50 percent.
05:18
But you know you're pretty close,
05:19
like I tell my kids all the time.
05:20
You're pretty close to a mouse.
05:22
There's not a big difference there.
05:23
That is a perfect segue.
05:25
Okay, so Steve has had some fantastic news lately.
05:29
I'm the mouse. (all laughing)
05:33
[Anne] Look at your DNA and see.
05:34
That's right, what defines a biologist,
05:36
it's a person who just really, really, really hates mice
05:39
and wants to torture them, right.
05:40
[Stephen] Mm-hmm. Exactly.
05:42
So you've been working on this cell atlas
05:44
which is this massive effort to just,
05:47
we don't even know how many different types of cells
05:49
there are in our bodies, and so it's a huge undertaking
05:53
and just last week you had some big news on that front.
05:55
We did, so the BioHub published
05:57
our mouse cell atlas in Nature and it was
06:01
the first whole organism atlas of cell types.
06:04
We're kind of on this brink of a great
06:07
technological revolution in cell biology
06:09
and all the tools of next generation sequencing
06:12
that revolutionized genetics over the last decade
06:15
are now poised to revolutionize cell biology.
06:17
And in particular we can ask questions about
06:21
what's the fundamental identity of each cell,
06:23
what's the plasticity in how they change
06:25
and what's their role in health and disease.
06:28
And the tools for measuring that have been
06:32
just maturing very nicely and the BioHub
06:35
provided an opportunity to kind of create
06:37
a nexus of technology and science
06:39
to take on a big project like that
06:41
and the mouse cell atlas, we call it Tabula Muris,
06:46
involved both the Stanford and UCSF medical schools.
06:49
About 15 different faculty groups
06:51
each of whom were experts at the various organs
06:56
that we analyzed and it was a great, great collaboration
06:59
in the bay area science community here.
07:02
And so what did you discover.
07:03
What's the big takeaway about the mouse cell atlas
07:05
that can tell us about the human.
07:07
Well it's a resource that you can use alongside
07:10
with the genome reference and so we're gonna be
07:14
figuring out different ways to use it
07:16
over the course of the next decade for sure.
07:17
But initial things are, what are the genes
07:21
that define cell type, and we now understand that
07:23
in 20 different tissues and organs
07:24
in a way that we didn't before.
07:26
Which are the genes we can use to reprogram
07:30
cell type and change identity.
07:31
We now have a list of hypotheses for those.
07:34
If you've got a new drug target you're interested in,
07:37
where is it expressed and where might you find toxicity.
07:39
People already using the atlas
07:41
to ask those kinds of questions.
07:43
There'll be many, many uses.
07:44
That's incredibly exciting.
07:45
So there's the famous Moore's law graph
07:49
which slices across your plane, like a very nice line
07:54
and then there's the cost of sequencing
07:57
that drops vertiginously in 2007
08:01
and that's kind of what set off
08:02
what I'm calling the neo-biological revolution,
08:05
which is all this innovation happening
08:07
in our bodies and in our medicine
08:08
in the way we think about what it means to be human.
08:12
So you have a list of technologies that have been
08:15
developed in the last 12 to 18 months
08:19
that made the cell atlas work possible.
08:21
Can you just talk about what those are?
08:22
Well I'd say it's like that old Hollywood chestnut.
08:26
It takes 20 years to become an overnight success.
08:29
It's a little bit like that with the cell atlas.
08:31
That these things have been brewing for quite some time
08:33
and it's just kind of all come together
08:35
over the past couple years in a nice way.
08:36
But it's building on a lot of work by a lot of groups.
08:39
My research and that of others.
08:40
But the abilities to enumerate
08:44
and count up every RNA molecule in a cell,
08:46
which is kind of the fundamental basis of this first atlas,
08:50
that depends on the sequencing technologies
08:51
that were many years in development.
08:53
It depends on microfluidic tools to isolate the cells
08:56
and break them open and capture those molecules.
08:58
And it depends on real advances
09:01
in computation and data infrastructure.
09:03
And that all kind of coalesced
09:04
very nicely in the past couple years.
09:07
So one of the things that Steve
09:09
has been such an innovator in,
09:10
is in non-invasive diagnostics testing
09:14
and so you were just referring to the microfluidics.
09:16
Can you talk about the liquid biopsy work
09:18
that you've been doing?
09:19
Sure.
09:20
So when I first became a parent
09:24
the doctor suggested an invasive biopsy
09:26
called amniocentesis for my wife and our unborn daughter.
09:29
And you know those are sort of scary things
09:32
that have negative health consequences
09:35
in the sense that there's a real health risk
09:36
for both the mom the baby to try to get that needle in there
09:39
and get some of the fetal cells.
09:40
And so that kind of stuck in my mind
09:42
as not a good thing and eventually my own research
09:46
turned toward addressing that question
09:48
and we developed the first, and it's a blood test
09:51
so it's non-invasive in the sense of not having any risk.
09:55
That has essentially replaced amniocentesis now.
09:57
Something like three million women
09:59
a year get some version of the test.
10:00
Amniocentesis rates have plunged 70 to 80 percent
10:03
and it's really kind of retiring.
10:07
Just completely transformational.
10:08
Yeah, I mean that was the scariest part
10:09
of the whole pregnancy.
10:10
I had no fears about the rest of the pregnancy.
10:12
It was that invasive procedure
10:14
that could have changed everything.
10:16
And the ideas we used to develop that
10:18
have now opened all kinds of other interesting doors
10:21
in medicine, which is this notion of,
10:23
there's many cases in medicine
10:24
where doctors are doing an invasive biopsy and it's risky
10:28
and difficult and so forth and the same ideas
10:30
and technologies can be applied in many other situations,
10:33
such as organ transplant recipients,
10:36
infectious disease, cancer, metabolic disease.
10:39
It's kind of going on and on.
10:41
So it's been, I really didn't expect
10:44
how many different directions it would take my research in.
10:47
So Anne you've been saying that you've been intending
10:51
to get involved in therapeutics all along,
10:53
that that was always your big idea there.
10:57
Well, I think it was that the idea of a company,
11:01
from what I learned when I was working on Wall Street.
11:03
I worked on Wall Street for 10 years
11:04
in health care investing and I realized
11:06
that there's a lot of, all of us can relate to
11:09
the experience of going to five different doctors
11:11
and getting five different opinions on
11:13
especially if it's something that's not well-treated.
11:17
And so I realized like one of the issues
11:19
there is, is there's not enough data.
11:21
And so I really look at 23andMe as a data company.
11:23
There's a great article years back about Target,
11:27
and they were able to identify a girl who was pregnant.
11:31
And I was like, Target knows exactly
11:33
when I walk in the door they know if I'm going left
11:35
or if I'm going right, and to the toys are to the right
11:37
so they definitely know I'm going to the right.
11:39
Like they know exactly our behavior
11:41
and I said like why isn't like that
11:43
when I go into my doctors office
11:44
that they can't have that same kind of predictions about me.
11:47
And I think part of the issue here
11:48
is there's not enough data.
11:49
And so 23andMe is really more of a data company
11:53
about amassing how much actually,
11:55
how many of our customers want to consent.
11:57
What is all the relevant information about yourself
11:59
in terms of diseases, lifestyle and can we actually
12:03
start to make those discoveries that are either
12:04
going to translate into a therapeutic
12:07
or in some ways like what's going to be most interesting
12:09
for all of us in the room is actually how do you prevent.
12:12
So something like Alzheimer's.
12:13
20 percent of the population is genetically
12:15
at increased risk for Alzheimer's.
12:17
And one of the issues I have in health care
12:20
is that wellness and prevention is not a reimbursable event.
12:24
So if we find we know how to potentially prevents
12:27
type 2 diabetes in those individuals,
12:29
but that's not well reimbursed versus
12:31
it's really well reimbursed to treat type 2 diabetes.
12:34
So I think it's one thing that's really interesting
12:36
with 23andMe is we're really well set up
12:38
to do drug discovery and that's always been a mission
12:41
to help people benefit, because people do get diseases.
12:43
But also really importantly is to help people understand
12:46
what is that balance between your genes and environment,
12:49
and what is it that you can change in your environment
12:51
that's going to help you impact the risk factors
12:55
that you're born with, and say, can you actually
12:57
have your best chance of not developing that disease.
12:59
So I really look as like, we start as data
13:02
and we're really about how can we help
13:04
our customers live to be healthy at 100.
13:06
And for some people it's going to be
13:08
that you need a medication, and for some people
13:10
you're gonna be able to take enough changes
13:12
in your lifestyle so you can actually be
13:15
not a well-medicated hospitalized 100 year old.
13:18
But actually a thriving yoga-doing 100 year old.
13:22
And so that's our goal.
13:24
So your vision of precision medicine
13:26
it depends on continuing to collect more data, right.
13:29
So you have five million profiles now.
13:32
80 percent of those people have agreed
13:34
to share their data for research.
13:36
Correct.
13:37
I think that precision medicine,
13:40
I have this piece, again when I was in Wall Street
13:42
back in 2000, the head of research and development of Roche.
13:46
She wrote this, in my mind it was a seminal piece,
13:49
it was right after the human genome had been sequenced.
13:52
It was in 2000, it was like the coming revolution
13:54
that's gonna come with personalized medicine.
13:57
And I remember thinking five, six years later,
14:01
there was nothing, like it wasn't moving forward.
14:04
And that's kind of when I decided to start 23andMe.
14:06
That if we want to see genetics happen, that the customer,
14:11
all of us, like we're going to be the ones who drive it.
14:13
We're gonna be the ones.
14:14
'Cause there's a lot of parts of the medical system
14:17
that are not necessarily supportive of personalized care.
14:20
Like even just think about our prevention recommendations.
14:24
Everyone should get a mammogram at 50.
14:27
Everyone gets a colonoscopy at 50.
14:29
[Stephen] Like, the reality-- Not everybody.
14:30
What did you say? Not everybody.
14:31
Not everybody, but that's the guideline.
14:33
I'm holding out on that one, all right.
14:36
You don't want to do it live? You can do it
14:38
live next year. That is gonna be
14:39
the next non-invasive test.
14:40
That would be amazing.
14:41
I'm keeping an eye on those very closely, all right.
14:43
(speakers laugh and audience applauds)
14:44
You got me working.
14:47
Well and to that point Steve, so what what can you do?
14:49
What other tools can you develop
14:51
for us that can drive us to.
14:53
(Steve and Anne laughs)
14:55
Well, you know it's a horrifying but true statistic
14:58
that 50 percent of all doctors graduate
15:01
in the bottom half of their class.
15:04
And you know that speaks to a larger issue
15:06
about equity in health care and access
15:08
and I think the things that Anne just talked about
15:11
are going to help address that.
15:12
The ability to share data transparently
15:14
across the world and best practices
15:17
and all this information, the fruits of
15:18
the genome revolution is gonna be very important.
15:21
But I also think there's many other ways
15:23
that that can come to fruition.
15:25
I'd mentioned the liquid biopsies
15:28
as a replacement for invasive biopsies.
15:30
It takes a lot of skill for a doctor to do those biopsies
15:32
and frankly the better doctors are in places with
15:36
the best hospitals and there's large parts of the country
15:38
and the world that don't have access to that
15:39
and to the extent that you can, and a key example
15:43
is for the organ transplant recipients.
15:45
If you get a harder kidney transplant, you're gonna do
15:47
a lot better if you're in the Bay Area than
15:50
if you're in the Midwest where there's not a major hospital.
15:53
Health by zip code.
15:54
Exactly, and so to the extent that
15:56
you can ameliorate that and level the field,
15:58
because blood can be drawn anywhere
16:00
and sent to a lab to be tested.
16:02
To the extent that you can provide
16:03
that kind of high-value information to people more broadly
16:06
it's a big step towards creating more equity in healthcare
16:09
which I think is a big part of
16:11
precision medicine from my perspective.
16:13
So obviously we're all very excited
16:15
about disruption in the health care space.
16:18
And the disruptors in chief at Amazon and JPMorgan Chase
16:24
and Berkshire Hathaway have now anointed Atul Gawande,
16:28
who's been a big deep thinker in this space
16:30
for a long time, to basically reinvent
16:33
what health care can look like going forward.
16:36
And everybody groused when he was
16:37
first appointed because he wasn't qualified.
16:39
He didn't have hospital administration experience.
16:42
He hadn't worked with insurance companies.
16:43
It's like yeah, that's the whole point.
16:45
So my question for you guys is,
16:47
what advice do you have for that effort for him.
16:52
Look, I think it's a great experiment that they're doing.
16:56
I was looking at this graph the other day,
16:58
this world of statistics thing
16:59
that looked at life expectancy versus amount of money spent
17:04
per person on health care for different countries.
17:06
And they had it for each country over time,
17:09
maybe you've seen this graph.
17:10
And in all of the developed world,
17:12
it kinda starts slow and it's increasing.
17:15
The more money they spend,
17:17
the higher life expectancy and then it plateaus.
17:20
And over the past few years it's plateaued
17:22
so they're spending more money
17:23
but not getting more life expectancy out.
17:26
And most of the developed world is on one curve there.
17:29
The US is a huge outlier.
17:31
We spend a factor of three to five more per individual
17:36
and get substantially lower life expectancy.
17:38
So, I don't know how you change
17:42
the system incrementally and I think bigger experiments
17:44
like that are an opportunity to try to
17:47
get us on the curve with the rest of the world.
17:49
I think one of the things I've always
17:51
found challenging in health care is that most health care,
17:56
especially at the private insurance level,
17:58
like it's slightly different in Medicare,
18:01
is everything's about, I'm willing to pay for something,
18:03
but it has to have a three year return on investment.
18:07
So if on average insurance companies
18:09
look at individuals and say I don't want,
18:12
so for instance most insurance companies
18:13
do not pay for people to get carrier status testing.
18:16
So things like cystic fibrosis, until you're pregnant.
18:20
Which is kind of unfortunate because then you're pregnant
18:22
and then you're potentially are carrying a child
18:23
with a condition and then you have worse decisions.
18:25
But the reason why they don't do that,
18:27
they don't offer it before, is because they don't want to
18:29
pay for the cost until they really have to.
18:32
Because they don't know are they going to benefit
18:34
from that cost versus someone else potentially,
18:36
'cause on average people switch insurers every three years.
18:40
So one of the biggest issues is that
18:42
decisions are made based on this
18:44
rolling three year time of when are you going to
18:48
actually see a return on your investment.
18:49
And that really distorts any kind of
18:52
long-term investment in your health.
18:54
Like one thing I often think about
18:55
is we're most aligned with life insurance
18:58
because if your life insurer actually care about
19:00
if you can be healthy at 100 that's better for you.
19:04
But your actual medical insurance,
19:06
it's really thinking about what do I have to pay for now.
19:09
It's a little bit like the game hot potato.
19:11
Like what do I have to pay for now
19:12
because someone else might be able to pay for that.
19:14
And I think that because this initiative
19:17
is so clearly focused on what can you really change
19:20
that's meaningful in health care
19:22
without necessarily the perspective.
19:23
Like there's lots of cost but it's that mindset of
19:27
what is actually the right long-term investment
19:29
in individuals and that's, again,
19:31
one of those issues that prevention
19:32
doesn't pay in the short term, it's a long-term investment.
19:35
So can they make those decisions
19:37
and is this the kind of company and initiative
19:39
that can afford to take on those initiatives
19:41
to really invest in the long-term outcome
19:44
of these individuals that they're covering.
19:46
And maybe scientific literacy is what's required
19:48
to get people to actually make that leap, right.
19:52
So the more you understand the amount of time it takes
19:54
for the science to develop and to be deployed.
19:56
I think scientific literacy is in some ways
19:58
is so different here because I think
19:59
in health care it's very practical.
20:01
It's very much about, I'm going,
20:05
health care in some ways is about
20:07
what you do every single day.
20:08
It's the sum of all of your actions.
20:10
I think scientific literacy is in some ways
20:12
a whole separate discussion about,
20:14
like I tell my therapeutics team.
20:16
Like our biggest competitor, the number one person
20:19
who's out there pulling eyeballs away
20:21
from us is Gwyneth Paltrow with Goop.
20:24
Like that's the reality is like people are fascinated
20:27
by things like Goop, and you see the controversy
20:30
that's around there so how do you distinguish between
20:34
Richard Scheller and you who's like well-known,
20:36
established scientists, and Gwyneth and other celebrities
20:40
who are coming up with theories.
20:42
So that to me is like the scientific literacy
20:44
and in some ways it starts like one thing
20:46
that we have to do is learn how to teach people
20:49
to ask questions to their physicians
20:51
and other groups so that they're aware.
20:53
Like we need a whole group whether it's in science
20:56
or it's in medicine about encouraging people
20:58
to have this responsibility, you are in charge of yourself
21:01
and you're totally capable of learning about yourself.
21:04
And that nothing scientifically is outside of your realm.
21:07
There's a lot of big words but it's just big words.
21:09
That's right.
21:10
Yeah, no I couldn't agree more.
21:11
I mean I really do feel like we have an obligation
21:13
to kind of help people understand that.
21:16
And you know my good friend Rob Phillips
21:18
has a great line about it.
21:19
He says, science is not an a la carte buffet.
21:22
You don't get to pick one thing over the other,
21:24
it's sort of what he means, so if you want to have airplanes
21:29
that get you around the world and cell phones
21:32
and the internet, you've also got to take
21:34
everything else that comes with science.
21:35
Evolution, climate change, vaccines, right.
21:40
It's all the same.
21:42
It's so good.
21:43
It's the same in intellectual foundation.
21:45
'Cause it's gonna happen either way.
21:46
[Anne] Yes.
21:47
And on that note we have to end.
21:49
But thank you so much.
21:50
[Anne] Thank you so much. Thank you.
21:52
(audience applauds)
More from wired.com
WANNA BUILD A ROBOT? NASA to Give Away a Mountain of Its Code
WANNA BUILD A ROBOT? NASA to Give Away a Mountain of Its Code
MONUMENT VALLEY An iPad Game to Make M.C. Escher Drool
MONUMENT VALLEY An iPad Game to Make M.C. Escher Drool
FRIEND REQUEST DENIED Classic Paintings, for Millennials
FRIEND REQUEST DENIED Classic Paintings, for Millennials
WELCOME TO SILICON VALLEY The Satire of Our Dreams With Mike Judge
WELCOME TO SILICON VALLEY The Satire of Our Dreams With Mike Judge
A BATTLE TO THE DEATH Uber vs. Lyft: The $500 Million Battle
A BATTLE TO THE DEATH Uber vs. Lyft: The $500 Million Battle
COBRA iRAD 230 Dodge Cops With This Radar Detector
COBRA iRAD 230 Dodge Cops With This Radar Detector