ABOUT THE SPEAKER
David Baker - Computational biologist
David Baker designs new biomolecules (proteins) from first principles to address 21st-century challenges in health and technology.

Why you should listen

David Baker is fascinated by biological self-organization. For example: How does the information stored in DNA translate into the intricate world of proteins and cells? The DNA code was solved more than 50 years ago, but the protein folding code has remained one of biology's greatest challenges. Starting 20 years ago, Baker's research team began using computers to model the structures of proteins. His work has advanced to the point where he can now not only predict the shape of natural proteins but also design completely new ones. In recent years, he's designed new experimental cancer therapies, vaccines, nanomaterials and more. He believes that the emerging field of protein design will fundamentally change how people make medicines, materials and more around the world. Now that the protein folding code is solved, the sky's the limit.

Baker is a Professor of Biochemistry and the Director of the Institute for Protein Design at the University of Washington in Seattle. He's also an Investigator at the Howard Hughes Medical Institute and Adjunct Professor of Genome Sciences, Bioengineering, Chemical Engineering, Computer Science, and Physics at the UW. With his colleagues, he developed the Rosetta Commons, the Rosetta@Home project and Foldit, a science video game. He has also launched more than ten companies that are seeking to bring designed proteins into the real world.

More profile about the speaker
David Baker | Speaker | TED.com
TED2019

David Baker: 5 challenges we could solve by designing new proteins

Filmed:
1,781,320 views

Proteins are remarkable molecular machines: they digest your food, fire your neurons, power your immune system and so much more. What if we could design new ones, with functions never before seen in nature? In this remarkable glimpse of the future, David Baker shares how his team at the Institute for Protein Design is creating entirely new proteins from scratch -- and shows how they could help us tackle five massive challenges facing humanity. (This ambitious plan is a part of the Audacious Project, TED's initiative to inspire and fund global change.)
- Computational biologist
David Baker designs new biomolecules (proteins) from first principles to address 21st-century challenges in health and technology. Full bio

Double-click the English transcript below to play the video.

00:12
I'm going to tell you about the most
amazing machines in the world
0
982
4059
00:17
and what we can now do with them.
1
5065
1768
00:19
Proteins,
2
7396
1163
00:20
some of which you see inside a cell here,
3
8583
2250
00:22
carry out essentially all the important
functions in our bodies.
4
10857
3459
00:26
Proteins digest your food,
5
14972
1879
00:28
contract your muscles,
6
16875
1706
00:30
fire your neurons
7
18605
1577
00:32
and power your immune system.
8
20206
1616
00:34
Everything that happens in biology --
9
22400
1976
00:36
almost --
10
24400
1151
00:37
happens because of proteins.
11
25575
1412
00:39
Proteins are linear chains
of building blocks called amino acids.
12
27698
4075
00:44
Nature uses an alphabet of 20 amino acids,
13
32366
3233
00:47
some of which have names
you may have heard of.
14
35623
2275
00:50
In this picture, for scale,
each bump is an atom.
15
38921
3542
00:55
Chemical forces between the amino acids
cause these long stringy molecules
16
43351
4644
01:00
to fold up into unique,
three-dimensional structures.
17
48019
3461
01:03
The folding process,
18
51937
1340
01:05
while it looks random,
19
53301
1434
01:06
is in fact very precise.
20
54759
1963
01:08
Each protein folds
to its characteristic shape each time,
21
56746
4397
01:13
and the folding process
takes just a fraction of a second.
22
61167
3388
01:18
And it's the shapes of proteins
23
66029
1844
01:19
which enable them to carry out
their remarkable biological functions.
24
67897
3970
01:24
For example,
25
72520
1151
01:25
hemoglobin has a shape
in the lungs perfectly suited
26
73695
3508
01:29
for binding a molecule of oxygen.
27
77227
1987
01:31
When hemoglobin moves to your muscle,
28
79759
1892
01:33
the shape changes slightly
29
81675
1932
01:35
and the oxygen comes out.
30
83631
2191
01:39
The shapes of proteins,
31
87494
1366
01:40
and hence their remarkable functions,
32
88884
2213
01:43
are completely specified by the sequence
of amino acids in the protein chain.
33
91121
5778
01:49
In this picture, each letter
on top is an amino acid.
34
97331
3941
01:54
Where do these sequences come from?
35
102860
1837
01:57
The genes in your genome
specify the amino acid sequences
36
105586
4824
02:02
of your proteins.
37
110434
1398
02:03
Each gene encodes the amino acid
sequence of a single protein.
38
111856
3738
02:09
The translation between
these amino acid sequences
39
117515
3802
02:13
and the structures
and functions of proteins
40
121341
2458
02:15
is known as the protein folding problem.
41
123823
2057
02:18
It's a very hard problem
42
126439
1545
02:20
because there's so many different
shapes a protein can adopt.
43
128008
3180
02:24
Because of this complexity,
44
132073
1645
02:25
humans have only been able
to harness the power of proteins
45
133742
2937
02:28
by making very small changes
to the amino acid sequences
46
136703
3468
02:32
of the proteins we've found in nature.
47
140195
2091
02:34
This is similar to the process
that our Stone Age ancestors used
48
142835
3858
02:38
to make tools and other implements
from the sticks and stones
49
146717
3359
02:42
that we found in the world around us.
50
150100
2003
02:45
But humans did not learn to fly
by modifying birds.
51
153226
5024
02:50
(Laughter)
52
158790
2017
02:52
Instead, scientists, inspired by birds,
uncovered the principles of aerodynamics.
53
160831
6310
02:59
Engineers then used those principles
to design custom flying machines.
54
167165
4395
03:04
In a similar way,
55
172195
1245
03:05
we've been working for a number of years
56
173464
1942
03:07
to uncover the fundamental
principles of protein folding
57
175430
3269
03:10
and encoding those principles
in the computer program called Rosetta.
58
178723
4059
03:15
We made a breakthrough in recent years.
59
183742
2513
03:19
We can now design completely new proteins
from scratch on the computer.
60
187029
4459
03:24
Once we've designed the new protein,
61
192396
2068
03:27
we encode its amino acid sequence
in a synthetic gene.
62
195242
3903
03:31
We have to make a synthetic gene
63
199656
1888
03:33
because since the protein
is completely new,
64
201568
2251
03:35
there's no gene in any organism on earth
which currently exists that encodes it.
65
203843
4762
03:41
Our advances in understanding
protein folding
66
209697
4187
03:45
and how to design proteins,
67
213908
1722
03:47
coupled with the decreasing cost
of gene synthesis
68
215654
3628
03:51
and the Moore's law increase
in computing power,
69
219306
3499
03:54
now enable us to design
tens of thousands of new proteins,
70
222829
4736
03:59
with new shapes and new functions,
71
227589
2339
04:01
on the computer,
72
229952
1513
04:03
and encode each one of those
in a synthetic gene.
73
231489
3915
04:08
Once we have those synthetic genes,
74
236248
1668
04:09
we put them into bacteria
75
237940
1545
04:11
to program them to make
these brand-new proteins.
76
239509
3305
04:15
We then extract the proteins
77
243197
2073
04:17
and determine whether they function
as we designed them to
78
245294
3436
04:20
and whether they're safe.
79
248754
1411
04:23
It's exciting to be able
to make new proteins,
80
251867
2465
04:26
because despite the diversity in nature,
81
254356
2496
04:28
evolution has only sampled a tiny fraction
of the total number of proteins possible.
82
256876
6092
04:35
I told you that nature uses
an alphabet of 20 amino acids,
83
263572
3495
04:39
and a typical protein is a chain
of about 100 amino acids,
84
267091
4449
04:43
so the total number of possibilities
is 20 times 20 times 20, 100 times,
85
271564
5552
04:49
which is a number on the order
of 10 to the 130th power,
86
277140
3817
04:52
which is enormously more
than the total number of proteins
87
280981
3812
04:56
which have existed
since life on earth began.
88
284817
2416
04:59
And it's this unimaginably large space
89
287990
2691
05:02
we can now explore
using computational protein design.
90
290705
3530
05:07
Now the proteins that exist on earth
91
295747
2369
05:10
evolved to solve the problems
faced by natural evolution.
92
298140
3993
05:14
For example, replicating the genome.
93
302705
2353
05:18
But we face new challenges today.
94
306128
2284
05:20
We live longer, so new
diseases are important.
95
308436
2737
05:23
We're heating up and polluting the planet,
96
311197
2215
05:25
so we face a whole host
of ecological challenges.
97
313436
3558
05:29
If we had a million years to wait,
98
317977
1808
05:31
new proteins might evolve
to solve those challenges.
99
319809
3208
05:35
But we don't have
millions of years to wait.
100
323787
2059
05:38
Instead, with computational
protein design,
101
326488
2871
05:41
we can design new proteins
to address these challenges today.
102
329383
4439
05:47
Our audacious idea is to bring
biology out of the Stone Age
103
335693
4450
05:52
through technological revolution
in protein design.
104
340167
2975
05:56
We've already shown
that we can design new proteins
105
344113
2864
05:59
with new shapes and functions.
106
347001
1683
06:01
For example, vaccines work
by stimulating your immune system
107
349174
4308
06:05
to make a strong response
against a pathogen.
108
353506
3122
06:09
To make better vaccines,
109
357698
1551
06:11
we've designed protein particles
110
359273
2302
06:13
to which we can fuse
proteins from pathogens,
111
361599
3587
06:17
like this blue protein here,
from the respiratory virus RSV.
112
365210
4334
06:22
To make vaccine candidates
113
370131
1730
06:23
that are literally bristling
with the viral protein,
114
371885
3663
06:27
we find that such vaccine candidates
115
375572
2570
06:30
produce a much stronger
immune response to the virus
116
378166
3302
06:33
than any previous vaccines
that have been tested.
117
381492
2703
06:36
This is important because RSV
is currently one of the leading causes
118
384648
3850
06:40
of infant mortality worldwide.
119
388522
2229
06:44
We've also designed new proteins
to break down gluten in your stomach
120
392414
3963
06:48
for celiac disease
121
396401
1597
06:50
and other proteins to stimulate
your immune system to fight cancer.
122
398022
4376
06:55
These advances are the beginning
of the protein design revolution.
123
403338
3939
07:00
We've been inspired by a previous
technological revolution:
124
408850
3190
07:04
the digital revolution,
125
412064
1345
07:05
which took place in large part
due to advances in one place,
126
413433
5125
07:10
Bell Laboratories.
127
418582
1272
07:12
Bell Labs was a place with an open,
collaborative environment,
128
420337
3294
07:15
and was able to attract top talent
from around the world.
129
423655
3183
07:19
And this led to a remarkable
string of innovations --
130
427418
3442
07:22
the transistor, the laser,
satellite communication
131
430884
4191
07:27
and the foundations of the internet.
132
435099
1726
07:29
Our goal is to build
the Bell Laboratories of protein design.
133
437761
3841
07:34
We are seeking to attract
talented scientists from around the world
134
442076
3515
07:37
to accelerate the protein
design revolution,
135
445615
2935
07:40
and we'll be focusing
on five grand challenges.
136
448574
4088
07:46
First, by taking proteins from flu strains
from around the world
137
454136
5597
07:51
and putting them on top
of the designed protein particles
138
459757
3554
07:55
I showed you earlier,
139
463335
1667
07:57
we aim to make a universal flu vaccine,
140
465026
3390
08:00
one shot of which gives a lifetime
of protection against the flu.
141
468440
3951
08:05
The ability to design --
142
473356
1612
08:06
(Applause)
143
474992
5224
08:12
The ability to design
new vaccines on the computer
144
480240
3068
08:15
is important both to protect
against natural flu epidemics
145
483332
5308
08:20
and, in addition, intentional
acts of bioterrorism.
146
488664
3480
08:25
Second, we're going far beyond
nature's limited alphabet
147
493272
3290
08:28
of just 20 amino acids
148
496586
1711
08:30
to design new therapeutic candidates
for conditions such as chronic pain,
149
498321
4735
08:35
using an alphabet
of thousands of amino acids.
150
503080
2631
08:38
Third, we're building
advanced delivery vehicles
151
506602
3813
08:42
to target existing medications
exactly where they need to go in the body.
152
510439
4164
08:47
For example, chemotherapy to a tumor
153
515226
2649
08:49
or gene therapies to the tissue
where gene repair needs to take place.
154
517899
4303
08:55
Fourth, we're designing smart therapeutics
that can do calculations within the body
155
523000
6532
09:01
and go far beyond current medicines,
156
529556
2214
09:03
which are really blunt instruments.
157
531794
2264
09:06
For example, to target a small
subset of immune cells
158
534082
4349
09:10
responsible for an autoimmune disorder,
159
538455
2081
09:12
and distinguish them from the vast
majority of healthy immune cells.
160
540560
3458
09:16
Finally, inspired by remarkable
biological materials
161
544899
3412
09:20
such as silk, abalone shell,
tooth and others,
162
548335
5108
09:25
we're designing new
protein-based materials
163
553467
2884
09:28
to address challenges in energy
and ecological issues.
164
556375
4163
09:33
To do all this,
we're growing our institute.
165
561558
2845
09:36
We seek to attract energetic,
talented and diverse scientists
166
564768
5599
09:42
from around the world,
at all career stages,
167
570391
3080
09:45
to join us.
168
573495
1150
09:47
You can also participate
in the protein design revolution
169
575304
3303
09:50
through our online
folding and design game, "Foldit."
170
578631
3744
09:55
And through our distributed
computing project, Rosetta@home,
171
583214
3851
09:59
which you can join from your laptop
or your Android smartphone.
172
587089
3731
10:04
Making the world a better place
through protein design is my life's work.
173
592547
3967
10:08
I'm so excited about
what we can do together.
174
596996
2278
10:11
I hope you'll join us,
175
599583
1470
10:13
and thank you.
176
601077
1158
10:14
(Applause and cheers)
177
602259
4455

▲Back to top

ABOUT THE SPEAKER
David Baker - Computational biologist
David Baker designs new biomolecules (proteins) from first principles to address 21st-century challenges in health and technology.

Why you should listen

David Baker is fascinated by biological self-organization. For example: How does the information stored in DNA translate into the intricate world of proteins and cells? The DNA code was solved more than 50 years ago, but the protein folding code has remained one of biology's greatest challenges. Starting 20 years ago, Baker's research team began using computers to model the structures of proteins. His work has advanced to the point where he can now not only predict the shape of natural proteins but also design completely new ones. In recent years, he's designed new experimental cancer therapies, vaccines, nanomaterials and more. He believes that the emerging field of protein design will fundamentally change how people make medicines, materials and more around the world. Now that the protein folding code is solved, the sky's the limit.

Baker is a Professor of Biochemistry and the Director of the Institute for Protein Design at the University of Washington in Seattle. He's also an Investigator at the Howard Hughes Medical Institute and Adjunct Professor of Genome Sciences, Bioengineering, Chemical Engineering, Computer Science, and Physics at the UW. With his colleagues, he developed the Rosetta Commons, the Rosetta@Home project and Foldit, a science video game. He has also launched more than ten companies that are seeking to bring designed proteins into the real world.

More profile about the speaker
David Baker | Speaker | TED.com