WEBVTT

99:59:59.999 --> 99:59:59.999
The second half of the last century

99:59:59.999 --> 99:59:59.999
was completely defined
by a technological revolution:

99:59:59.999 --> 99:59:59.999
the software revolution.

99:59:59.999 --> 99:59:59.999
The ability to program electrons
on a material called silicon

99:59:59.999 --> 99:59:59.999
made possible technologies,
companies and industries

99:59:59.999 --> 99:59:59.999
that were at one point
unimaginable to many of us,

99:59:59.999 --> 99:59:59.999
but which have now fundamentally changed
the way the world works.

99:59:59.999 --> 99:59:59.999
The first half of this century, though,

99:59:59.999 --> 99:59:59.999
is going to be transformed
by a new software revolution:

99:59:59.999 --> 99:59:59.999
the living software revolution.

99:59:59.999 --> 99:59:59.999
And this will be powered by the ability
to program biochemistry

99:59:59.999 --> 99:59:59.999
on a material called biology,

99:59:59.999 --> 99:59:59.999
and doing so will enable us to harness
the properties of biology

99:59:59.999 --> 99:59:59.999
to generate new kinds of therapies,

99:59:59.999 --> 99:59:59.999
to repair damaged tissue,

99:59:59.999 --> 99:59:59.999
to reprogram faulty cells

99:59:59.999 --> 99:59:59.999
or even build programmable
operating systems out of biochemistry.

99:59:59.999 --> 99:59:59.999
If we can realize this,
and we do need to realize it,

99:59:59.999 --> 99:59:59.999
its impact will be so enormous

99:59:59.999 --> 99:59:59.999
that it will make the first
software revolution pale in comparison,

99:59:59.999 --> 99:59:59.999
and that's because living software
would transform the entirety of medicine,

99:59:59.999 --> 99:59:59.999
agriculture, and energy,

99:59:59.999 --> 99:59:59.999
and these are sectors that dwarf
those dominated by IT.

99:59:59.999 --> 99:59:59.999
Imagine programmable plants
that fix nitrogen more effectively

99:59:59.999 --> 99:59:59.999
or resist emerging fungal pathogens,

99:59:59.999 --> 99:59:59.999
or even programming crops
to be perennial rather than annual

99:59:59.999 --> 99:59:59.999
so you could double
your crop yields each year.

99:59:59.999 --> 99:59:59.999
That would transform agriculture

99:59:59.999 --> 99:59:59.999
and how we'll keep our growing
and global population fed.

99:59:59.999 --> 99:59:59.999
Or imagine programmable immunity,

99:59:59.999 --> 99:59:59.999
designing and harnessing molecular devices
that guide your immune system

99:59:59.999 --> 99:59:59.999
to detect, eradicate,
or even prevent disease.

99:59:59.999 --> 99:59:59.999
This would transform medicine

99:59:59.999 --> 99:59:59.999
and how we'll keep our growing
and aging population healthy.

99:59:59.999 --> 99:59:59.999
We already have many of the tools
that will make living software a reality.

99:59:59.999 --> 99:59:59.999
We can precisely edit genes with CRISPR.

99:59:59.999 --> 99:59:59.999
We can rewrite the genetic code
one base at a time.

99:59:59.999 --> 99:59:59.999
We can even build functioning
synthetic circuits out of DNA.

99:59:59.999 --> 99:59:59.999
But figuring out how and when
to wield these tools

99:59:59.999 --> 99:59:59.999
is still a process of trial and error.

99:59:59.999 --> 99:59:59.999
It needs deep expertise,

99:59:59.999 --> 99:59:59.999
years of specialization,

99:59:59.999 --> 99:59:59.999
and experimental protocols
are difficult to discover

99:59:59.999 --> 99:59:59.999
and all too often difficult to reproduce.

99:59:59.999 --> 99:59:59.999
And, you know, we have a tendency
in biology to focus a lot on the parts

99:59:59.999 --> 99:59:59.999
but we all know that something like flying

99:59:59.999 --> 99:59:59.999
wouldn't be understood
by only studying feathers.

99:59:59.999 --> 99:59:59.999
So programming biology is not yet
as simple as programming your computer,

99:59:59.999 --> 99:59:59.999
and then to make matters worse

99:59:59.999 --> 99:59:59.999
living systems largely bear no resemblance
to the engineered systems

99:59:59.999 --> 99:59:59.999
that you and I program every day.

99:59:59.999 --> 99:59:59.999
In contrast to engineered systems,
living systems self-generate,

99:59:59.999 --> 99:59:59.999
they self-organize,

99:59:59.999 --> 99:59:59.999
they operate at molecular scales,

99:59:59.999 --> 99:59:59.999
and these molecular-level interactions

99:59:59.999 --> 99:59:59.999
lead generally to robust
macro-scale output.

99:59:59.999 --> 99:59:59.999
They can even self-repair.

99:59:59.999 --> 99:59:59.999
Consider, for example,
the humble household plant,

99:59:59.999 --> 99:59:59.999
like that one that sat
on your mantlepiece at home

99:59:59.999 --> 99:59:59.999
that you keep forgetting to water.

99:59:59.999 --> 99:59:59.999
Every day, despite your neglect,
that plant has to wake up

99:59:59.999 --> 99:59:59.999
and figure out how
to allocate its resources.

99:59:59.999 --> 99:59:59.999
Will it grow, photosynthesize,
produce seeds, or flower?

99:59:59.999 --> 99:59:59.999
And that's a decision that has to be made
at the level of the whole organism.

99:59:59.999 --> 99:59:59.999
But a plant doesn't have a brain
to figure all of that out.

99:59:59.999 --> 99:59:59.999
It has to make do
with the cells on its leaves.

99:59:59.999 --> 99:59:59.999
They have to respond to the environment

99:59:59.999 --> 99:59:59.999
and make the decisions
that affect the whole plant.

99:59:59.999 --> 99:59:59.999
So somehow there must be
a program running inside these cells,

99:59:59.999 --> 99:59:59.999
a program that responds
to input signals and cues

99:59:59.999 --> 99:59:59.999
and shapes what that cell will do.

99:59:59.999 --> 99:59:59.999
And then those programs must operate

99:59:59.999 --> 99:59:59.999
in a distributed way
across individual cells

99:59:59.999 --> 99:59:59.999
so that they can coordinate
and that plant can grow and flourish.

99:59:59.999 --> 99:59:59.999
If we could understand
these biological programs,

99:59:59.999 --> 99:59:59.999
if we could understand
biological computation,

99:59:59.999 --> 99:59:59.999
it would transform our ability
to understand how and why

99:59:59.999 --> 99:59:59.999
cells do what they do.

99:59:59.999 --> 99:59:59.999
Because, if we understood these programs,

99:59:59.999 --> 99:59:59.999
we could debug them
when things go wrong,

99:59:59.999 --> 99:59:59.999
or we could learn from them how
to design the kind of synthetic circuits

99:59:59.999 --> 99:59:59.999
that truly exploit

99:59:59.999 --> 99:59:59.999
the computational power of biochemistry.

NOTE Paragraph

99:59:59.999 --> 99:59:59.999
My passion about this idea
led me to a career in research

99:59:59.999 --> 99:59:59.999
at the interface of maths,
computer science, and biology,

99:59:59.999 --> 99:59:59.999
and in my work I focus on the concept
of biology as computation.

99:59:59.999 --> 99:59:59.999
And that means asking
what the cells compute,

99:59:59.999 --> 99:59:59.999
and how can we uncover
these biological programs?

99:59:59.999 --> 99:59:59.999
And I started to ask these questions
together with some brilliant collaborators

99:59:59.999 --> 99:59:59.999
at Microsoft Research
and the University of Cambridge,

99:59:59.999 --> 99:59:59.999
where together we wanted to understand

99:59:59.999 --> 99:59:59.999
the biological program
running inside a unique type of cell:

99:59:59.999 --> 99:59:59.999
an embryonic stem cell.

99:59:59.999 --> 99:59:59.999
These cells are unique
because they're totally naïve.

99:59:59.999 --> 99:59:59.999
They can become anything they want:

99:59:59.999 --> 99:59:59.999
a brain cell, a heart cell,
a bone cell, a lung cell,

99:59:59.999 --> 99:59:59.999
any adult cell type.

99:59:59.999 --> 99:59:59.999
This naïvety, it sets them apart,

99:59:59.999 --> 99:59:59.999
but it also ignited the imagination
of the scientific community,

99:59:59.999 --> 99:59:59.999
who realized, if we could
tap into that potential,

99:59:59.999 --> 99:59:59.999
we would have a powerful
tool for medicine.

99:59:59.999 --> 99:59:59.999
If we could figure out
how these cells make the decision

99:59:59.999 --> 99:59:59.999
to become one cell type or another,

99:59:59.999 --> 99:59:59.999
we might be able to harness them

99:59:59.999 --> 99:59:59.999
to generate cells that we need
to repair diseased or damaged tissue.

99:59:59.999 --> 99:59:59.999
But realizing that vision
is not without its challenges,

99:59:59.999 --> 99:59:59.999
not least because these particular cells,

99:59:59.999 --> 99:59:59.999
they emerge just six days
after conception,

99:59:59.999 --> 99:59:59.999
and then within a day or so, they're gone.

99:59:59.999 --> 99:59:59.999
They have set off down the different paths

99:59:59.999 --> 99:59:59.999
that form all the structures
and organs of your adult body.

99:59:59.999 --> 99:59:59.999
But it turns out that cell fates
are a lot more plastic

99:59:59.999 --> 99:59:59.999
than we might have imagined.

99:59:59.999 --> 99:59:59.999
About 13 years ago, some scientists
showed something truly revolutionary.

99:59:59.999 --> 99:59:59.999
By inserting just a handful of genes
into an adult cell,

99:59:59.999 --> 99:59:59.999
like one of your skin cells,

99:59:59.999 --> 99:59:59.999
you can transform that cell
back to the naïve state.

99:59:59.999 --> 99:59:59.999
And it's a process that's actually
known as reprogramming,

99:59:59.999 --> 99:59:59.999
and it allows us to imagine
a kind of stem cell utopia,

99:59:59.999 --> 99:59:59.999
the ability to take a sample
of a patient's own cells,

99:59:59.999 --> 99:59:59.999
transform them back to the naïve state,

99:59:59.999 --> 99:59:59.999
and use those cells to make
whatever that patient might need,

99:59:59.999 --> 99:59:59.999
whether it's brain cells or heart cells.

99:59:59.999 --> 99:59:59.999
But over the last decade or so,

99:59:59.999 --> 99:59:59.999
figuring out how to change cell fate,

99:59:59.999 --> 99:59:59.999
it's still a process of trial and error.

99:59:59.999 --> 99:59:59.999
Even in cases where we've uncovered
successful experimental protocols,

99:59:59.999 --> 99:59:59.999
they're still inefficient,

99:59:59.999 --> 99:59:59.999
and we lack a fundamental understanding
of how and why they work.

99:59:59.999 --> 99:59:59.999
If you've figured out how to change
a stem cell into a heart cell,

99:59:59.999 --> 99:59:59.999
that hasn't got any way of telling you
how to change a stem cell

99:59:59.999 --> 99:59:59.999
into a brain cell.

99:59:59.999 --> 99:59:59.999
So we wanted to understand
the biological program

99:59:59.999 --> 99:59:59.999
running inside an embryonic stem cell,

99:59:59.999 --> 99:59:59.999
and understanding the computation
performed by a living system

99:59:59.999 --> 99:59:59.999
starts with asking
a devastatingly simple question:

99:59:59.999 --> 99:59:59.999
what is it that system actually has to do?

99:59:59.999 --> 99:59:59.999
Now, computer science actually has
a set of strategies for dealing

99:59:59.999 --> 99:59:59.999
with what it is that software
and hardware are meant to do.

99:59:59.999 --> 99:59:59.999
When you write a program,
you code a piece of software,

99:59:59.999 --> 99:59:59.999
you want that software to run correctly.

99:59:59.999 --> 99:59:59.999
You want performance, functionality.

99:59:59.999 --> 99:59:59.999
You want to prevent bugs.

99:59:59.999 --> 99:59:59.999
They can cost you a lot.

99:59:59.999 --> 99:59:59.999
So when a developer writes a program,

99:59:59.999 --> 99:59:59.999
they could write down
a set of specifications.

99:59:59.999 --> 99:59:59.999
These are what your program should do.

99:59:59.999 --> 99:59:59.999
Maybe it should compare
the size of two numbers,

99:59:59.999 --> 99:59:59.999
or order numbers by increasing size.

99:59:59.999 --> 99:59:59.999
Technology exists

99:59:59.999 --> 99:59:59.999
that allows us automatically to check

99:59:59.999 --> 99:59:59.999
whether our specifications are satisfied,

99:59:59.999 --> 99:59:59.999
whether that program
does what it should do.

99:59:59.999 --> 99:59:59.999
And so our idea
was that in the same way,

99:59:59.999 --> 99:59:59.999
experimental observations,
things we measure in the lab,

99:59:59.999 --> 99:59:59.999
they correspond to specifications

99:59:59.999 --> 99:59:59.999
of what the biological program should do.

99:59:59.999 --> 99:59:59.999
So we just needed to figure out a way

99:59:59.999 --> 99:59:59.999
to encode this new type of specification.

99:59:59.999 --> 99:59:59.999
So let's say you've been busy in the lab
and you've been measuring your genes

99:59:59.999 --> 99:59:59.999
and you've found that if Gene A is active,

99:59:59.999 --> 99:59:59.999
then Gene B or Gene C seems to be active.

99:59:59.999 --> 99:59:59.999
We can write that observation down
as a mathematical expression

99:59:59.999 --> 99:59:59.999
if we can use the language of logic.

99:59:59.999 --> 99:59:59.999
If A, then B or C.

99:59:59.999 --> 99:59:59.999
Now, this is a very simple example, OK.

99:59:59.999 --> 99:59:59.999
It's just to illustrate the point.

99:59:59.999 --> 99:59:59.999
We can encode truly rich expressions

99:59:59.999 --> 99:59:59.999
that actually capture the behavior
of multiple genes or proteins over time

99:59:59.999 --> 99:59:59.999
across multiple different experiments.

99:59:59.999 --> 99:59:59.999
And so by translating our observations

99:59:59.999 --> 99:59:59.999
into mathematical expression in this way,

99:59:59.999 --> 99:59:59.999
it becomes possible to test
whether or not those observations

99:59:59.999 --> 99:59:59.999
can emerge from a program
of genetic interactions.

NOTE Paragraph

99:59:59.999 --> 99:59:59.999
And we developed a tool to do just this.

99:59:59.999 --> 99:59:59.999
We were able to use this tool
to encode observations

99:59:59.999 --> 99:59:59.999
as mathematical expressions,

99:59:59.999 --> 99:59:59.999
and then that tool would allow us
to uncover the genetic program

99:59:59.999 --> 99:59:59.999
that could explain them all.

99:59:59.999 --> 99:59:59.999
And we then apply this approach

99:59:59.999 --> 99:59:59.999
to uncover the genetic program
running inside embryonic stem cells

99:59:59.999 --> 99:59:59.999
to see if we could understand
how to induce that naïve state.

99:59:59.999 --> 99:59:59.999
And this tool was actually built

99:59:59.999 --> 99:59:59.999
on a ?? that's deployed
routinely around the world

99:59:59.999 --> 99:59:59.999
for conventional software verification.

99:59:59.999 --> 99:59:59.999
So we started with a set
of nearly 50 different specifications

99:59:59.999 --> 99:59:59.999
that we generated from experimental
observations of embryonic stem cells,

99:59:59.999 --> 99:59:59.999
and by encoding these
observations in this tool,

99:59:59.999 --> 99:59:59.999
we were able to uncover
the first molecular program

99:59:59.999 --> 99:59:59.999
that could explain all of them.

99:59:59.999 --> 99:59:59.999
Now, that's kind of a feat
in and of itself, right?

99:59:59.999 --> 99:59:59.999
Being able to reconcile
all of these different observations

99:59:59.999 --> 99:59:59.999
is not the kind of thing
you can do on the back of an envelope,

99:59:59.999 --> 99:59:59.999
even if you have a really big envelope.

99:59:59.999 --> 99:59:59.999
Because we'd got
this kind of understanding,

99:59:59.999 --> 99:59:59.999
we could go one step further.

99:59:59.999 --> 99:59:59.999
We could use this program to predict
what this cell might do

99:59:59.999 --> 99:59:59.999
in conditions we hadn't yet tested.

99:59:59.999 --> 99:59:59.999
We could probe the program in ??.

99:59:59.999 --> 99:59:59.999
And so we did just that:

99:59:59.999 --> 99:59:59.999
we generated predictions
that we tested in the lab,

99:59:59.999 --> 99:59:59.999
and we found that this program
was highly predictive.

99:59:59.999 --> 99:59:59.999
It told us how we could
accelerate progress

99:59:59.999 --> 99:59:59.999
back to the naïve state
quickly and efficiently.

99:59:59.999 --> 99:59:59.999
It told us which genes
to target to do that,

99:59:59.999 --> 99:59:59.999
which genes might even
hinder that process.

99:59:59.999 --> 99:59:59.999
We even found the program predicted
the order in which genes would switch on.

99:59:59.999 --> 99:59:59.999
So this approach really allowed us
to uncover the dynamics

99:59:59.999 --> 99:59:59.999
of what the cells are doing.

99:59:59.999 --> 99:59:59.999
Now, what we've developed,
it's not a method

99:59:59.999 --> 99:59:59.999
that's specific to stem cell biology.

99:59:59.999 --> 99:59:59.999
Rather, it allows us to make sense

99:59:59.999 --> 99:59:59.999
of the computation
being carried out by the cell

99:59:59.999 --> 99:59:59.999
in the context of genetic interactions.

99:59:59.999 --> 99:59:59.999
So really, it's just one building block.

99:59:59.999 --> 99:59:59.999
The field urgently needs
to develop new approaches

99:59:59.999 --> 99:59:59.999
to understand biological
computation more broadly

99:59:59.999 --> 99:59:59.999
and at different levels,

99:59:59.999 --> 99:59:59.999
from DNA right through
to the flow of information between cells.

99:59:59.999 --> 99:59:59.999
Only this kind of
transformative understanding

99:59:59.999 --> 99:59:59.999
will enable us to harness biology
in ways that are predictable and reliable.

99:59:59.999 --> 99:59:59.999
But to program biology,
we will also need to develop

99:59:59.999 --> 99:59:59.999
the kinds of tools and languages

99:59:59.999 --> 99:59:59.999
that allow both experimentalists
and computational scientists

99:59:59.999 --> 99:59:59.999
to design biological function

99:59:59.999 --> 99:59:59.999
and have those designs compile down
to the machine code of the cell,

99:59:59.999 --> 99:59:59.999
its biochemistry,

99:59:59.999 --> 99:59:59.999
so that we could then
build those structures.

99:59:59.999 --> 99:59:59.999
Now, that's something akin
to a living software compiler,

99:59:59.999 --> 99:59:59.999
and I'm proud to be
part of a team at Microsoft

99:59:59.999 --> 99:59:59.999
that's working to develop one.

99:59:59.999 --> 99:59:59.999
Though to say it's a grand challenge
is kind of an understatement,

99:59:59.999 --> 99:59:59.999
but if it's realized,

99:59:59.999 --> 99:59:59.999
it would be the final bridge
between software and wetware.

99:59:59.999 --> 99:59:59.999
More broadly, though, programming biology
is only going to be possible

99:59:59.999 --> 99:59:59.999
if we can transform the field
into being truly interdisciplinary.

99:59:59.999 --> 99:59:59.999
It needs us to bridge
the physical and the life sciences,

99:59:59.999 --> 99:59:59.999
and scientists from
each of these disciplines

99:59:59.999 --> 99:59:59.999
need to be able to work together
with common languages

99:59:59.999 --> 99:59:59.999
and to have shared scientific questions.

NOTE Paragraph

99:59:59.999 --> 99:59:59.999
In the long term, it's worth remembering
that many of the giant software companies

99:59:59.999 --> 99:59:59.999
and the technology
that you and I work with every day

99:59:59.999 --> 99:59:59.999
could hardly have been imagined

99:59:59.999 --> 99:59:59.999
at the time we first started
programming on silicon microchips,

99:59:59.999 --> 99:59:59.999
and if we start now to think about
the potential for technology

99:59:59.999 --> 99:59:59.999
enabled by computational biology,

99:59:59.999 --> 99:59:59.999
we'll see some of the steps
that we need to take along the way

99:59:59.999 --> 99:59:59.999
to make that a reality.

99:59:59.999 --> 99:59:59.999
Now, there is the sobering thought
that this kind of technology

99:59:59.999 --> 99:59:59.999
could be open to misuse.

99:59:59.999 --> 99:59:59.999
If we're willing to talk
about the potential

99:59:59.999 --> 99:59:59.999
for programming immune cells,

99:59:59.999 --> 99:59:59.999
we should also be thinking
about the potential of bacteria

99:59:59.999 --> 99:59:59.999
engineered to evade them.

99:59:59.999 --> 99:59:59.999
There might be people willing to do that.

99:59:59.999 --> 99:59:59.999
Now, one reassuring thought in this

99:59:59.999 --> 99:59:59.999
is that, well, less so for the scientists,

99:59:59.999 --> 99:59:59.999
is that biology is
a fragile thing to work with.

99:59:59.999 --> 99:59:59.999
So programming biology
is not going to be something

99:59:59.999 --> 99:59:59.999
you'll be doing in your garden shed.

99:59:59.999 --> 99:59:59.999
But because we're at the outset of this,

99:59:59.999 --> 99:59:59.999
we can move forward
with our eyes wide open.

99:59:59.999 --> 99:59:59.999
We can ask the difficult
questions up front,

99:59:59.999 --> 99:59:59.999
we can put in place
the necessary safeguards,

99:59:59.999 --> 99:59:59.999
and as part of that,
we'll have to think about our ethics.

99:59:59.999 --> 99:59:59.999
We'll have to think about putting bounds

99:59:59.999 --> 99:59:59.999
on the implementation
of biological function.

99:59:59.999 --> 99:59:59.999
So as part of this research in bioethics
will have to be a priority.

99:59:59.999 --> 99:59:59.999
It can't be relegated to second place

99:59:59.999 --> 99:59:59.999
in the excitement
of scientific innovation.

99:59:59.999 --> 99:59:59.999
But the ultimate prize,

99:59:59.999 --> 99:59:59.999
the ultimate destination on this journey,

99:59:59.999 --> 99:59:59.999
would be breakthrough applications
and breakthrough industries

99:59:59.999 --> 99:59:59.999
in areas from agriculture and medicine
to energy and materials

99:59:59.999 --> 99:59:59.999
and even computing itself.

99:59:59.999 --> 99:59:59.999
Imagine, one day we could be powering
the planet sustainably

99:59:59.999 --> 99:59:59.999
on the ultimate green energy
if we could mimic something

99:59:59.999 --> 99:59:59.999
that plants figured out millennia ago:

99:59:59.999 --> 99:59:59.999
how to harness the Sun's energy
with an efficiency that is unparalleled

99:59:59.999 --> 99:59:59.999
by our current solar cells.

99:59:59.999 --> 99:59:59.999
If we understood that program
of quantum interactions

99:59:59.999 --> 99:59:59.999
that allow plants to absorb
sunlight so efficiently,

99:59:59.999 --> 99:59:59.999
we might be able to translate that
into building synthetic DNA circuits

99:59:59.999 --> 99:59:59.999
that offer the material
for better solar cells.

99:59:59.999 --> 99:59:59.999
There are teams and scientists working
on the fundamentals of this right now,

99:59:59.999 --> 99:59:59.999
so perhaps if it got the right attention
and the right investment,

99:59:59.999 --> 99:59:59.999
it could be realized in 10 or 15 years.

99:59:59.999 --> 99:59:59.999
So we are at the beginning
of a technological revolution.

99:59:59.999 --> 99:59:59.999
Understanding this ancient type
of biological computation

99:59:59.999 --> 99:59:59.999
is the critical first step,
and if we can realize this,

99:59:59.999 --> 99:59:59.999
we would enter in the era
of an operating system

99:59:59.999 --> 99:59:59.999
that runs living software.

99:59:59.999 --> 99:59:59.999
Thank you very much.

99:59:59.999 --> 99:59:59.999
(Applause)