0:00:00.964,0:00:03.509
The second half of the last century

0:00:03.509,0:00:07.608
was completely defined[br]by a technological revolution:

0:00:07.608,0:00:09.512
the software revolution.

0:00:09.512,0:00:14.145
The ability to program electrons[br]on a material called silicon

0:00:14.145,0:00:17.444
made possible technologies,[br]companies and industries

0:00:17.444,0:00:21.451
that were at one point[br]unimaginable to many of us,

0:00:21.451,0:00:25.895
but which have now fundamentally changed[br]the way the world works.

0:00:25.895,0:00:28.126
The first half of this century, though,

0:00:28.126,0:00:32.393
is going to be transformed[br]by a new software revolution:

0:00:32.393,0:00:35.137
the living software revolution.

0:00:35.137,0:00:39.172
And this will be powered by the ability[br]to program biochemistry

0:00:39.172,0:00:41.807
on a material called biology,

0:00:41.807,0:00:45.479
and doing so will enable us to harness[br]the properties of biology

0:00:45.479,0:00:48.417
to generate new kinds of therapies,

0:00:48.417,0:00:50.883
to repair damaged tissue,

0:00:50.883,0:00:53.127
to reprogram faulty cells

0:00:53.127,0:00:57.768
or even build programmable[br]operating systems out of biochemistry.

0:00:58.659,0:01:01.846
If we can realize this,[br]and we do need to realize it,

0:01:01.846,0:01:04.855
its impact will be so enormous

0:01:04.855,0:01:08.104
that it will make the first[br]software revolution pale in comparison,

0:01:08.104,0:01:12.282
and that's because living software[br]would transform the entirety of medicine,

0:01:12.282,0:01:14.475
agriculture, and energy,

0:01:14.475,0:01:18.292
and these are sectors that dwarf[br]those dominated by IT.

0:01:18.292,0:01:23.010
Imagine programmable plants[br]that fix nitrogen more effectively

0:01:23.010,0:01:26.084
or resist emerging fungal pathogens,

0:01:26.212,0:01:29.500
or even programming crops[br]to be perennial rather than annual

0:01:29.500,0:01:33.285
so you could double[br]your crop yields each year.

0:01:33.285,0:01:34.431
That would transform agriculture

0:01:34.431,0:01:39.000
and how we'll keep our growing[br]and global population fed.

0:01:39.000,0:01:42.013
Or imagine programmable immunity,

0:01:42.013,0:01:45.533
designing and harnessing molecular devices[br]that guide your immune system

0:01:45.533,0:01:49.130
to detect, eradicate,[br]or even prevent disease.

0:01:49.414,0:01:51.075
This would transform medicine

0:01:51.075,0:01:54.564
and how we'll keep our growing[br]and aging population healthy.

0:01:55.373,0:02:00.235
We already have many of the tools[br]that will make living software a reality.

0:02:00.235,0:02:02.844
We can precisely edit genes with CRISPR.

0:02:02.844,0:02:05.278
We can rewrite the genetic code[br]one base at a time.

0:02:05.278,0:02:10.266
We can even build functioning[br]synthetic circuits out of DNA.

0:02:10.266,0:02:13.956
But figuring out how and when[br]to wield these tools

0:02:13.956,0:02:15.513
is still a process of trial and error.

0:02:15.513,0:02:17.988
It needs deep expertise,

0:02:17.988,0:02:19.138
years of specialization,

0:02:19.138,0:02:21.969
and experimental protocols[br]are difficult to discover

0:02:21.969,0:02:24.487
and all too often difficult to reproduce.

0:02:25.502,0:02:28.986
And, you know, we have a tendency[br]in biology to focus a lot on the parts

0:02:28.986,0:02:32.391
but we all know that something like flying

0:02:32.391,0:02:34.997
wouldn't be understood[br]by only studying feathers.

0:02:34.997,0:02:40.054
So programming biology is not yet[br]as simple as programming your computer,

0:02:40.054,0:02:41.511
and then to make matters worse

0:02:41.511,0:02:45.501
living systems largely bear no resemblance[br]to the engineered systems

0:02:45.501,0:02:47.977
that you and I program every day.

0:02:47.977,0:02:52.017
In contrast to engineered systems,[br]living systems self-generate,

0:02:52.017,0:02:53.556
they self-organize,

0:02:53.556,0:02:55.266
they operate at molecular scales,

0:02:55.266,0:02:57.595
and these molecular-level interactions

0:02:57.595,0:03:00.884
lead generally to robust[br]macro-scale output.

0:03:00.884,0:03:03.929
They can even self-repair.

0:03:04.256,0:03:07.526
Consider, for example,[br]the humble household plant,

0:03:07.526,0:03:09.642
like that one that sat[br]on your mantlepiece at home

0:03:09.642,0:03:11.537
that you keep forgetting to water.

0:03:11.537,0:03:15.388
Every day, despite your neglect,[br]that plant has to wake up

0:03:15.388,0:03:18.322
and figure out how[br]to allocate its resources.

0:03:18.322,0:03:21.996
Will it grow, photosynthesize,[br]produce seeds, or flower?

0:03:21.996,0:03:25.748
And that's a decision that has to be made[br]at the level of the whole organism.

0:03:25.973,0:03:29.611
But a plant doesn't have a brain[br]to figure all of that out.

0:03:29.611,0:03:32.185
It has to make do[br]with the cells on its leaves.

0:03:32.185,0:03:34.084
They have to respond to the environment

0:03:34.084,0:03:37.090
and make the decisions[br]that affect the whole plant.

0:03:37.090,0:03:40.748
So somehow there must be[br]a program running inside these cells,

0:03:40.748,0:03:44.307
a program that responds[br]to input signals and cues

0:03:44.307,0:03:45.865
and shapes what that cell will do.

0:03:45.865,0:03:47.714
And then those programs must operate

0:03:47.714,0:03:50.311
in a distributed way[br]across individual cells

0:03:50.311,0:03:54.434
so that they can coordinate[br]and that plant can grow and flourish.

0:03:55.074,0:03:59.356
If we could understand[br]these biological programs,

0:03:59.356,0:04:02.427
if we could understand[br]biological computation,

0:04:02.427,0:04:06.878
it would transform our ability[br]to understand how and why

0:04:06.878,0:04:08.540
cells do what they do.

0:04:08.540,0:04:11.168
Because, if we understood these programs,

0:04:11.168,0:04:12.524
we could debug them[br]when things go wrong,

0:04:12.524,0:04:16.059
or we could learn from them how[br]to design the kind of synthetic circuits

0:04:16.059,0:04:19.459
that truly exploit

0:04:19.459,0:04:21.978
the computational power of biochemistry.

0:04:21.978,0:04:26.068
My passion about this idea[br]led me to a career in research

0:04:26.068,0:04:29.742
at the interface of maths,[br]computer science, and biology,

0:04:29.742,0:04:34.596
and in my work I focus on the concept[br]of biology as computation.

0:04:34.596,0:04:37.689
And that means asking[br]what the cells compute,

0:04:37.689,0:04:41.839
and how can we uncover[br]these biological programs?

0:04:42.031,0:04:45.444
And I started to ask these questions[br]together with some brilliant collaborators

0:04:45.444,0:04:48.601
at Microsoft Research[br]and the University of Cambridge,

0:04:48.601,0:04:51.333
where together we wanted to understand

0:04:51.333,0:04:54.890
the biological program[br]running inside a unique type of cell:

0:04:54.890,0:04:56.989
an embryonic stem cell.

0:04:57.280,0:05:00.463
These cells are unique[br]because they're totally naïve.

0:05:00.463,0:05:02.911
They can become anything they want:

0:05:02.911,0:05:06.433
a brain cell, a heart cell,[br]a bone cell, a lung cell,

0:05:06.433,0:05:07.359
any adult cell type.

0:05:07.359,0:05:09.033
This naïvety, it sets them apart,

0:05:09.033,0:05:11.780
but it also ignited the imagination[br]of the scientific community,

0:05:11.780,0:05:15.432
who realized, if we could[br]tap into that potential,

0:05:15.432,0:05:18.235
we would have a powerful[br]tool for medicine.

0:05:18.235,0:05:20.511
If we could figure out[br]how these cells make the decision

0:05:20.511,0:05:22.646
to become one cell type or another,

0:05:22.646,0:05:24.740
we might be able to harness them

0:05:24.740,0:05:30.175
to generate cells that we need[br]to repair diseased or damaged tissue.

0:05:30.175,0:05:33.707
But realizing that vision[br]is not without its challenges,

0:05:33.707,0:05:35.941
not least because these particular cells,

0:05:35.941,0:05:39.847
they emerge just six days[br]after conception,

0:05:39.847,0:05:41.216
and then within a day or so, they're gone.

0:05:41.216,0:05:43.972
They have set off down the different paths

0:05:43.972,0:05:47.046
that form all the structures[br]and organs of your adult body.

0:05:47.046,0:05:50.562
But it turns out that cell fates[br]are a lot more plastic

0:05:50.562,0:05:52.653
than we might have imagined.

0:05:52.653,0:05:57.171
About 13 years ago, some scientists[br]showed something truly revolutionary.

0:05:57.171,0:06:00.480
By inserting just a handful of genes[br]into an adult cell,

0:06:00.480,0:06:02.300
like one of your skin cells,

0:06:02.300,0:06:07.813
you can transform that cell[br]back to the naïve state.

0:06:07.813,0:06:11.807
And it's a process that's actually[br]known as reprogramming,

0:06:11.807,0:06:14.414
and it allows us to imagine[br]a kind of stem cell utopia,

0:06:14.414,0:06:17.847
the ability to take a sample[br]of a patient's own cells,

0:06:17.847,0:06:20.859
transform them back to the naïve state,

0:06:20.859,0:06:24.157
and use those cells to make[br]whatever that patient might need,

0:06:24.157,0:06:25.651
whether it's brain cells or heart cells.

0:06:25.651,0:06:29.400
But over the last decade or so,

0:06:29.400,0:06:31.731
figuring out how to change cell fate,

0:06:31.731,0:06:34.094
it's still a process of trial and error.

0:06:34.094,0:06:37.917
Even in cases where we've uncovered[br]successful experimental protocols,

0:06:37.917,0:06:40.787
they're still inefficient,

0:06:40.787,0:06:44.857
and we lack a fundamental understanding[br]of how and why they work.

0:06:44.999,0:06:47.437
If you've figured out how to change[br]a stem cell into a heart cell,

0:06:47.437,0:06:49.553
that hasn't got any way of telling you[br]how to change a stem cell

0:06:49.553,0:06:52.871
into a brain cell.

0:06:52.871,0:06:55.739
So we wanted to understand[br]the biological program

0:06:55.739,0:06:58.209
running inside an embryonic stem cell,

0:06:58.209,0:07:01.738
and understanding the computation[br]performed by a living system

0:07:01.738,0:07:05.921
starts with asking[br]a devastatingly simple question:

0:07:05.921,0:07:10.100
what is it that system actually has to do?

0:07:10.100,0:07:13.082
Now, computer science actually has[br]a set of strategies for dealing

0:07:13.082,0:07:16.762
with what it is that software[br]and hardware are meant to do.

0:07:16.762,0:07:19.593
When you write a program,[br]you code a piece of software,

0:07:19.593,0:07:23.117
you want that software to run correctly.

0:07:23.117,0:07:23.594
You want performance, functionality.

0:07:23.594,0:07:24.558
You want to prevent bugs.

0:07:24.558,0:07:26.248
They can cost you a lot.

0:07:26.248,0:07:27.896
So when a developer writes a program,

0:07:27.896,0:07:30.508
they could write down[br]a set of specifications.

0:07:30.508,0:07:32.334
These are what your program should do.

0:07:32.334,0:07:34.788
Maybe it should compare[br]the size of two numbers,

0:07:34.788,0:07:36.684
or order numbers by increasing size.

0:07:36.684,0:07:39.316
Technology exists

0:07:39.316,0:07:41.867
that allows us automatically to check

0:07:41.867,0:07:45.086
whether our specifications are satisfied,

0:07:45.086,0:07:47.465
whether that program[br]does what it should do.

0:07:47.465,0:07:50.395
And so our idea[br]was that in the same way,

0:07:50.395,0:07:53.800
experimental observations,[br]things we measure in the lab,

0:07:53.800,0:07:56.282
they correspond to specifications

0:07:56.282,0:07:58.988
of what the biological program should do.

0:07:59.150,0:08:01.298
So we just needed to figure out a way

0:08:01.298,0:08:04.213
to encode this new type of specification.

0:08:04.729,0:08:07.796
So let's say you've been busy in the lab[br]and you've been measuring your genes

0:08:07.796,0:08:10.685
and you've found that if Gene A is active,

0:08:10.685,0:08:14.073
then Gene B or Gene C seems to be active.

0:08:14.623,0:08:18.541
We can write that observation down[br]as a mathematical expression

0:08:18.541,0:08:21.591
if we can use the language of logic.

0:08:21.591,0:08:24.242
If A, then B or C.

0:08:24.242,0:08:27.009
Now, this is a very simple example, OK.

0:08:27.009,0:08:28.627
It's just to illustrate the point.

0:08:28.627,0:08:31.757
We can encode truly rich expressions

0:08:31.757,0:08:35.357
that actually capture the behavior[br]of multiple genes or proteins over time

0:08:35.357,0:08:38.641
across multiple different experiments.

0:08:38.641,0:08:41.603
And so by translating our observations

0:08:41.603,0:08:43.988
into mathematical expression in this way,

0:08:43.988,0:08:47.541
it becomes possible to test[br]whether or not those observations

0:08:47.541,0:08:51.939
can emerge from a program[br]of genetic interactions.

0:08:52.262,0:08:54.841
And we developed a tool to do just this.

0:08:54.841,0:08:57.205
We were able to use this tool[br]to encode observations

0:08:57.205,0:08:59.181
as mathematical expressions,

0:08:59.181,0:09:03.120
and then that tool would allow us[br]to uncover the genetic program

0:09:03.120,0:09:05.704
that could explain them all.

0:09:05.704,0:09:08.914
And we then apply this approach

0:09:08.914,0:09:12.066
to uncover the genetic program[br]running inside embryonic stem cells

0:09:12.066,0:09:16.328
to see if we could understand[br]how to induce that naïve state.

0:09:16.328,0:09:19.011
And this tool was actually built

0:09:19.011,0:09:21.711
on a ?? that's deployed[br]routinely around the world

0:09:21.711,0:09:23.901
for conventional software verification.

0:09:23.901,0:09:27.592
So we started with a set[br]of nearly 50 different specifications

0:09:27.592,0:09:32.269
that we generated from experimental[br]observations of embryonic stem cells,

0:09:32.269,0:09:34.928
and by encoding these[br]observations in this tool,

0:09:34.928,0:09:38.563
we were able to uncover[br]the first molecular program

0:09:38.563,0:09:40.548
that could explain all of them.

0:09:40.548,0:09:42.632
Now, that's kind of a feat[br]in and of itself, right?

0:09:42.632,0:09:45.772
Being able to reconcile[br]all of these different observations

0:09:45.772,0:09:49.133
is not the kind of thing[br]you can do on the back of an envelope,

0:09:49.133,0:09:51.805
even if you have a really big envelope.

0:09:51.805,0:09:54.986
Because we'd got[br]this kind of understanding,

0:09:54.986,0:09:56.558
we could go one step further.

0:09:56.558,0:09:59.474
We could use this program to predict[br]what this cell might do

0:09:59.474,0:10:01.855
in conditions we hadn't yet tested.

0:10:01.855,0:10:04.614
We could probe the program in ??.

0:10:04.614,0:10:07.074
And so we did just that:

0:10:07.074,0:10:09.001
we generated predictions[br]that we tested in the lab,

0:10:09.001,0:10:12.306
and we found that this program[br]was highly predictive.

0:10:12.306,0:10:15.510
It told us how we could[br]accelerate progress

0:10:15.510,0:10:18.250
back to the naïve state[br]quickly and efficiently.

0:10:18.250,0:10:20.733
It told us which genes[br]to target to do that,

0:10:20.733,0:10:23.281
which genes might even[br]hinder that process.

0:10:23.281,0:10:28.841
We even found the program predicted[br]the order in which genes would switch on.

0:10:29.187,0:10:32.414
So this approach really allowed us[br]to uncover the dynamics

0:10:32.414,0:10:35.009
of what the cells are doing.

0:10:35.265,0:10:39.476
Now, what we've developed,[br]it's not a method

0:10:39.476,0:10:40.286
that's specific to stem cell biology.

0:10:40.286,0:10:42.832
Rather, it allows us to make sense

0:10:42.832,0:10:43.940
of the computation[br]being carried out by the cell

0:10:43.940,0:10:46.751
in the context of genetic interactions.

0:10:46.751,0:10:49.271
So really, it's just one building block.

0:10:49.271,0:10:51.897
The field urgently needs[br]to develop new approaches

0:10:51.897,0:10:54.697
to understand biological[br]computation more broadly

0:10:54.697,0:10:57.146
and at different levels,

0:10:57.146,0:10:59.930
from DNA right through[br]to the flow of information between cells.

0:10:59.930,0:11:03.288
Only this kind of[br]transformative understanding

0:11:03.288,0:11:07.515
will enable us to harness biology[br]in ways that are predictable and reliable.

0:11:09.205,0:11:12.395
But to program biology,[br]we will also need to develop

0:11:12.395,0:11:14.926
the kinds of tools and languages

0:11:14.926,0:11:17.759
that allow both experimentalists[br]and computational scientists

0:11:17.759,0:11:20.260
to design biological function

0:11:20.439,0:11:23.773
and have those designs compile down[br]to the machine code of the cell,

0:11:23.773,0:11:25.371
its biochemistry,

0:11:25.371,0:11:27.228
so that we could then[br]build those structures.

0:11:27.228,0:11:31.073
Now, that's something akin[br]to a living software compiler,

0:11:31.073,0:11:33.622
and I'm proud to be[br]part of a team at Microsoft

0:11:33.622,0:11:35.572
that's working to develop one.

0:11:35.572,0:11:39.712
Though to say it's a grand challenge[br]is kind of an understatement,

0:11:39.712,0:11:41.939
but if it's realized,

0:11:41.939,0:11:44.356
it would be the final bridge[br]between software and wetware.

0:11:44.794,0:11:48.315
More broadly, though, programming biology[br]is only going to be possible

0:11:48.315,0:11:52.924
if we can transform the field[br]into being truly interdisciplinary.

0:11:52.924,0:11:56.244
It needs us to bridge[br]the physical and the life sciences,

0:11:56.244,0:11:58.288
and scientists from[br]each of these disciplines

0:11:58.288,0:12:01.037
need to be able to work together[br]with common languages

0:12:01.037,0:12:03.923
and to have shared scientific questions.

0:12:04.884,0:12:08.561
In the long term, it's worth remembering[br]that many of the giant software companies

0:12:08.561,0:12:10.875
and the technology[br]that you and I work with every day

0:12:10.875,0:12:12.200
could hardly have been imagined

0:12:12.200,0:12:16.039
at the time we first started[br]programming on silicon microchips,

0:12:16.402,0:12:20.128
and if we start now to think about[br]the potential for technology

0:12:20.128,0:12:22.148
enabled by computational biology,

0:12:22.148,0:12:24.959
we'll see some of the steps[br]that we need to take along the way

0:12:24.959,0:12:27.132
to make that a reality.

0:12:27.132,0:12:30.582
Now, there is the sobering thought[br]that this kind of technology

0:12:30.582,0:12:32.356
could be open to misuse.

0:12:32.356,0:12:34.330
If we're willing to talk[br]about the potential

0:12:34.330,0:12:35.890
for programming immune cells,

0:12:35.890,0:12:39.215
we should also be thinking[br]about the potential of bacteria

0:12:39.215,0:12:41.200
engineered to evade them.

0:12:41.200,0:12:43.290
There might be people willing to do that.

0:12:43.290,0:12:45.575
Now, one reassuring thought in this

0:12:45.575,0:12:47.732
is that, well, less so for the scientists,

0:12:47.732,0:12:51.066
is that biology is[br]a fragile thing to work with.

0:12:51.066,0:12:53.100
So programming biology[br]is not going to be something

0:12:53.100,0:12:55.168
you'll be doing in your garden shed.

0:12:55.459,0:12:57.987
But because we're at the outset of this,

0:12:57.987,0:13:00.593
we can move forward[br]with our eyes wide open.

0:13:00.593,0:13:02.916
We can ask the difficult[br]questions up front,

0:13:02.916,0:13:05.972
we can put in place[br]the necessary safeguards,

0:13:05.972,0:13:08.919
and as part of that,[br]we'll have to think about our ethics.

0:13:08.919,0:13:10.820
We'll have to think about putting bounds

0:13:10.820,0:13:13.786
on the implementation[br]of biological function.

0:13:13.946,0:13:16.780
So as part of this research in bioethics[br]will have to be a priority.

0:13:16.780,0:13:18.271
It can't be relegated to second place

0:13:18.271,0:13:22.500
in the excitement[br]of scientific innovation.

0:13:24.035,0:13:25.235
But the ultimate prize,

0:13:25.235,0:13:26.483
the ultimate destination on this journey,

0:13:26.483,0:13:28.589
would be breakthrough applications[br]and breakthrough industries

0:13:28.589,0:13:31.619
in areas from agriculture and medicine[br]to energy and materials

0:13:31.619,0:13:35.733
and even computing itself.

0:13:36.490,0:13:40.334
Imagine, one day we could be powering[br]the planet sustainably

0:13:40.334,0:13:42.039
on the ultimate green energy[br]if we could mimic something

0:13:42.039,0:13:45.738
that plants figured out millennia ago:

0:13:45.738,0:13:49.251
how to harness the Sun's energy[br]with an efficiency that is unparalleled

0:13:49.251,0:13:51.648
by our current solar cells.

0:13:51.648,0:13:54.465
If we understood that program[br]of quantum interactions

0:13:54.465,0:13:57.936
that allow plants to absorb[br]sunlight so efficiently,

0:13:57.936,0:14:01.765
we might be able to translate that[br]into building synthetic DNA circuits

0:14:01.765,0:14:05.427
that offer the material[br]for better solar cells.

0:14:05.651,0:14:08.947
There are teams and scientists working[br]on the fundamentals of this right now,

0:14:08.947,0:14:11.771
so perhaps if it got the right attention[br]and the right investment,

0:14:11.771,0:14:15.268
it could be realized in 10 or 15 years.

0:14:15.712,0:14:18.947
So we are at the beginning[br]of a technological revolution.

0:14:19.171,0:14:22.656
Understanding this ancient type[br]of biological computation

0:14:22.656,0:14:24.850
is the critical first step,[br]and if we can realize this,

0:14:24.850,0:14:28.711
we would enter in the era[br]of an operating system

0:14:28.711,0:14:30.936
that runs living software.

0:14:30.936,0:14:32.691
Thank you very much.

0:14:32.691,0:14:34.510
(Applause)