TED2014
Stephen Friend: The hunt for "unexpected genetic heroes"
Filmed:
Readability: 4.3
1,017,016 views
What can we learn from people with the genetics to get sick — who don't? With most inherited diseases, only some family members will develop the disease, while others who carry the same genetic risks dodge it. Stephen Friend suggests we start studying those family members who stay healthy. Hear about the Resilience Project, a massive effort to collect genetic materials that may help decode inherited disorders.
Stephen Friend - Open-science advocate
Inspired by open-source software models, Sage Bionetworks co-founder Stephen Friend builds tools that facilitate research sharing on a massive and revolutionary scale. Full bio
Inspired by open-source software models, Sage Bionetworks co-founder Stephen Friend builds tools that facilitate research sharing on a massive and revolutionary scale. Full bio
Double-click the English transcript below to play the video.
00:12
Approximately 30 years ago,
0
602
2338
00:14
when I was in oncology at the Children's Hospital
1
2940
2693
00:17
in Philadelphia,
2
5633
1389
00:19
a father and a son walked into my office
3
7022
3154
00:22
and they both had their right eye missing,
4
10176
3144
00:25
and as I took the history, it became apparent
5
13320
2811
00:28
that the father and the son had a rare form
6
16131
2769
00:30
of inherited eye tumor, retinoblastoma,
7
18900
3542
00:34
and the father knew that he had passed that fate
8
22442
3114
00:37
on to his son.
9
25556
1875
00:39
That moment changed my life.
10
27431
2412
00:41
It propelled me to go on
11
29843
1904
00:43
and to co-lead a team that discovered
12
31747
3532
00:47
the first cancer susceptibility gene,
13
35279
3197
00:50
and in the intervening decades since then,
14
38476
2721
00:53
there has been literally a seismic shift
15
41197
3420
00:56
in our understanding of what goes on,
16
44617
2026
00:58
what genetic variations are sitting behind
17
46643
2888
01:01
various diseases.
18
49531
1559
01:03
In fact, for thousands of human traits,
19
51090
3384
01:06
a molecular basis that's known for that,
20
54474
2218
01:08
and for thousands of people, every day,
21
56692
3295
01:11
there's information that they gain
22
59987
2081
01:14
about the risk of going on to get this disease
23
62068
2442
01:16
or that disease.
24
64510
2226
01:18
At the same time, if you ask,
25
66736
2305
01:21
"Has that impacted the efficiency,
26
69041
2707
01:23
how we've been able to develop drugs?"
27
71748
2092
01:25
the answer is not really.
28
73840
1782
01:27
If you look at the cost of developing drugs,
29
75622
2330
01:29
how that's done, it basically hasn't budged that.
30
77952
3389
01:33
And so it's as if we have the power to diagnose
31
81341
4473
01:37
yet not the power to fully treat.
32
85814
2812
01:40
And there are two commonly given reasons
33
88626
2466
01:43
for why that happens.
34
91092
1468
01:44
One of them is it's early days.
35
92560
3472
01:48
We're just learning the words, the fragments,
36
96032
3590
01:51
the letters in the genetic code.
37
99622
1776
01:53
We don't know how to read the sentences.
38
101398
2155
01:55
We don't know how to follow the narrative.
39
103553
2570
01:58
The other reason given is that
40
106123
2479
02:00
most of those changes are a loss of function,
41
108602
2218
02:02
and it's actually really hard to develop drugs
42
110820
2925
02:05
that restore function.
43
113745
1915
02:07
But today, I want us to step back
44
115660
2182
02:09
and ask a more fundamental question,
45
117842
2028
02:11
and ask, "What happens if we're thinking
46
119870
2189
02:14
about this maybe in the wrong context?"
47
122059
2733
02:16
We do a lot of studying of those who are sick
48
124792
3159
02:19
and building up long lists
49
127951
2600
02:22
of altered components.
50
130551
3118
02:25
But maybe, if what we're trying to do
51
133669
2399
02:28
is to develop therapies for prevention,
52
136068
3222
02:31
maybe what we should be doing
53
139290
1553
02:32
is studying those who don't get sick.
54
140843
2382
02:35
Maybe we should be studying those
55
143225
2347
02:37
that are well.
56
145572
2175
02:39
A vast majority of those people
57
147747
1797
02:41
are not necessarily carrying a particular
58
149544
2336
02:43
genetic load or risk factor.
59
151880
1936
02:45
They're not going to help us.
60
153816
1984
02:47
There are going to be those individuals
61
155800
1599
02:49
who are carrying a potential future risk,
62
157399
2669
02:52
they're going to go on to get some symptom.
63
160068
1844
02:53
That's not what we're looking for.
64
161912
1788
02:55
What we're asking and looking for is,
65
163700
1848
02:57
are there a very few set of individuals
66
165548
2770
03:00
who are actually walking around
67
168318
2836
03:03
with the risk that normally would cause a disease,
68
171154
4019
03:07
but something in them, something hidden in them
69
175173
2963
03:10
is actually protective
70
178136
1834
03:11
and keeping them from exhibiting those symptoms?
71
179970
3175
03:15
If you're going to do a study
like that, you can imagine
like that, you can imagine
72
183145
2053
03:17
you'd like to look at lots and lots of people.
73
185198
2832
03:20
We'd have to go and have a pretty wide study,
74
188030
3292
03:23
and we realized that actually
75
191322
1735
03:25
one way to think of this is,
76
193057
1529
03:26
let us look at adults who are over 40 years of age,
77
194586
4277
03:30
and let's make sure that we look at those
78
198863
2970
03:33
who were healthy as kids.
79
201833
1480
03:35
They might have had individuals in their families
80
203313
2402
03:37
who had had a childhood disease,
81
205715
1812
03:39
but not necessarily.
82
207527
1506
03:41
And let's go and then screen those
83
209033
2767
03:43
to find those who are carrying genes
84
211800
1993
03:45
for childhood diseases.
85
213793
1678
03:47
Now, some of you, I can see you
86
215471
1564
03:49
putting your hands up going, "Uh, a little odd.
87
217035
3295
03:52
What's your evidence
88
220330
1417
03:53
that this could be feasible?"
89
221747
1662
03:55
I want to give you two examples.
90
223409
2064
03:57
The first comes from San Francisco.
91
225473
2948
04:00
It comes from the 1980s and the 1990s,
92
228421
2941
04:03
and you may know the story where
93
231362
2394
04:05
there were individuals who had very high levels
94
233756
2397
04:08
of the virus HIV.
95
236153
1268
04:09
They went on to get AIDS.
96
237421
2479
04:11
But there was a very small set of individuals
97
239900
2317
04:14
who also had very high levels of HIV.
98
242217
2968
04:17
They didn't get AIDS.
99
245185
1386
04:18
And astute clinicians tracked that down,
100
246571
2962
04:21
and what they found was
they were carrying mutations.
they were carrying mutations.
101
249533
3387
04:24
Notice, they were carrying mutations from birth
102
252920
3085
04:28
that were protective, that were protecting them
103
256005
2015
04:30
from going on to get AIDS.
104
258020
1641
04:31
You may also know that actually a line of therapy
105
259661
3165
04:34
has been coming along based on that fact.
106
262826
3120
04:37
Second example, more recent, is elegant work
107
265946
3224
04:41
done by Helen Hobbs,
108
269170
1403
04:42
who said, "I'm going to look at individuals
109
270573
2662
04:45
who have very high lipid levels,
110
273235
2716
04:47
and I'm going to try to find those people
111
275951
1939
04:49
with high lipid levels
112
277890
1802
04:51
who don't go on to get heart disease."
113
279692
2168
04:53
And again, what she found was
114
281860
2438
04:56
some of those individuals had mutations
115
284298
2560
04:58
that were protective from birth that kept them,
116
286858
2719
05:01
even though they had high lipid levels,
117
289577
1445
05:03
and you can see this is an interesting way
118
291022
3371
05:06
of thinking about how you could develop
119
294393
1961
05:08
preventive therapies.
120
296354
2260
05:10
The project that we're working on
121
298614
1944
05:12
is called "The Resilience Project:
122
300558
2462
05:15
A Search for Unexpected Heroes,"
123
303020
1400
05:16
because what we are interested in doing is saying,
124
304420
2490
05:18
can we find those rare individuals
125
306910
2648
05:21
who might have these hidden protective factors?
126
309558
4325
05:25
And in some ways, think of it as a decoder ring,
127
313883
2980
05:28
a sort of resilience decoder ring
128
316863
1926
05:30
that we're going to try to build.
129
318789
1632
05:32
We've realized that we should
do this in a systematic way,
do this in a systematic way,
130
320421
3849
05:36
so we've said, let's take every single
131
324270
2627
05:38
childhood inherited disease.
132
326897
1243
05:40
Let's take them all, and let's
pull them back a little bit
pull them back a little bit
133
328140
2564
05:42
by those that are known to have severe symptoms,
134
330704
3186
05:45
where the parents, the child,
135
333890
1920
05:47
those around them would know
136
335810
1050
05:48
that they'd gotten sick,
137
336860
1330
05:50
and let's go ahead and then frame them again
138
338190
3700
05:53
by those parts of the genes where we know
139
341890
2581
05:56
that there is a particular alteration
140
344471
2507
05:58
that is known to be highly penetrant
141
346978
2798
06:01
to cause that disease.
142
349776
2654
06:04
Where are we going to look?
143
352430
1228
06:05
Well, we could look locally. That makes sense.
144
353658
2488
06:08
But we began to think, maybe we should look
145
356146
2261
06:10
all over the world.
146
358407
1451
06:11
Maybe we should look not just here
147
359858
1653
06:13
but in remote places where their might be
148
361511
1960
06:15
a distinct genetic context,
149
363471
3030
06:18
there might be environmental factors
150
366501
1642
06:20
that protect people.
151
368143
1382
06:21
And let's look at a million individuals.
152
369525
4462
06:25
Now the reason why we think it's a good time
153
373987
2970
06:28
to do that now
154
376957
1072
06:30
is, in the last couple of years,
155
378029
1760
06:31
there's been a remarkable plummeting in the cost
156
379789
2588
06:34
to do this type of analysis,
157
382377
2235
06:36
this type of data generation,
158
384612
1739
06:38
to where it actually costs less to do
159
386351
2608
06:40
the data generation and analysis
160
388959
2194
06:43
than it does to do the sample
processing and the collection.
processing and the collection.
161
391153
3184
06:46
The other reason is that in the last five years,
162
394337
4304
06:50
there have been awesome tools,
163
398641
1964
06:52
things about network biology, systems biology,
164
400605
2662
06:55
that have come up that allow us to think
165
403267
1961
06:57
that maybe we could decipher
166
405228
1940
06:59
those positive outliers.
167
407168
2481
07:01
And as we went around talking to researchers
168
409649
2172
07:03
and institutions
169
411821
1904
07:05
and telling them about our story,
170
413725
1569
07:07
something happened.
171
415294
1667
07:08
They started saying, "This is interesting.
172
416961
2229
07:11
I would be glad to join your effort.
173
419190
3347
07:14
I would be willing to participate."
174
422537
1927
07:16
And they didn't say, "Where's the MTA?"
175
424464
2579
07:19
They didn't say, "Where is my authorship?"
176
427043
3293
07:22
They didn't say, "Is this data going
to be mine? Am I going to own it?"
to be mine? Am I going to own it?"
177
430336
4611
07:26
They basically said, "Let's work on this
178
434947
2279
07:29
in an open, crowd-sourced, team way
179
437226
2881
07:32
to do this decoding."
180
440107
3074
07:35
Six months ago, we locked down
181
443181
2515
07:37
the screening key for this decoder.
182
445696
3315
07:41
My co-lead, a brilliant scientist, Eric Schadt
183
449011
4578
07:45
at the Icahn Mount Sinai
School of Medicine in New York,
School of Medicine in New York,
184
453589
3306
07:48
and his team,
185
456895
1392
07:50
locked in that decoder key ring,
186
458287
2869
07:53
and we began looking for samples,
187
461156
2395
07:55
because what we realized is,
188
463551
1486
07:57
maybe we could just go and look
189
465037
1794
07:58
at some existing samples to
get some sense of feasibility.
get some sense of feasibility.
190
466831
3086
08:01
Maybe we could take two, three
percent of the project on,
percent of the project on,
191
469917
2577
08:04
and see if it was there.
192
472494
1417
08:05
And so we started asking people
193
473911
1998
08:07
such as Hakon at the Children's Hospital in Philadelphia.
194
475909
3537
08:11
We asked Leif up in Finland.
195
479446
2245
08:13
We talked to Anne Wojcicki at 23andMe,
196
481691
3673
08:17
and Wang Jun at BGI,
197
485364
1767
08:19
and again, something remarkable happened.
198
487131
2188
08:21
They said, "Huh,
199
489319
1809
08:23
not only do we have samples,
200
491128
1744
08:24
but often we've analyzed them,
201
492872
2196
08:27
and we would be glad to go into
202
495068
1487
08:28
our anonymized samples
203
496555
1403
08:29
and see if we could find those
204
497958
2062
08:32
that you're looking for."
205
500020
1163
08:33
And instead of being 20,000 or 30,000,
206
501183
2707
08:35
last month we passed one half million samples
207
503890
3152
08:39
that we've already analyzed.
208
507042
1905
08:40
So you must be going,
209
508947
1493
08:42
"Huh, did you find any unexpected heroes?"
210
510440
5625
08:48
And the answer is, we didn't find one or two.
211
516065
2583
08:50
We found dozens of these strong candidate
212
518648
3038
08:53
unexpected heroes.
213
521686
1729
08:55
So we think that the time is now
214
523415
2697
08:58
to launch the beta phase of this project
215
526112
2340
09:00
and actually start getting prospective individuals.
216
528452
3117
09:03
Basically all we need is information.
217
531569
3171
09:06
We need a swab of DNA
218
534740
1659
09:08
and a willingness to say, "What's inside me?
219
536399
3405
09:11
I'm willing to be re-contacted."
220
539804
3263
09:15
Most of us spend our lives,
221
543067
3791
09:18
when it comes to health and disease,
222
546858
1954
09:20
acting as if we're voyeurs.
223
548812
3080
09:23
We delegate the responsibility
224
551892
2337
09:26
for the understanding of our disease,
225
554229
2043
09:28
for the treatment of our disease,
226
556272
1872
09:30
to anointed experts.
227
558144
3536
09:33
In order for us to get this project to work,
228
561680
3340
09:37
we need individuals to step up
229
565020
2150
09:39
in a different role and to be engaged,
230
567170
3892
09:43
to realize this dream,
231
571062
2925
09:45
this open crowd-sourced project,
232
573987
3135
09:49
to find those unexpected heroes,
233
577122
3680
09:52
to evolve from the current concepts
234
580802
2660
09:55
of resources and constraints,
235
583462
2334
09:57
to design those preventive therapies,
236
585796
3251
10:01
and to extend it beyond childhood diseases,
237
589047
2773
10:03
to go all the way up to ways
238
591820
1577
10:05
that we could look at Alzheimer's or Parkinson's,
239
593397
3871
10:09
we're going to need us
240
597268
2262
10:11
to be looking inside ourselves and asking,
241
599530
3106
10:14
"What are our roles?
242
602636
2204
10:16
What are our genes?"
243
604840
1673
10:18
and looking within ourselves for information
244
606513
2785
10:21
we used to say we should go to the outside,
245
609298
2642
10:23
to experts,
246
611940
1208
10:25
and to be willing to share that with others.
247
613148
4052
10:29
Thank you very much.
248
617200
3558
10:32
(Applause)
249
620758
1815
ABOUT THE SPEAKER
Stephen Friend - Open-science advocateInspired by open-source software models, Sage Bionetworks co-founder Stephen Friend builds tools that facilitate research sharing on a massive and revolutionary scale.
Why you should listen
While working for Merck, Stephen Friend became frustrated by the slow pace at which big pharma created new treatments for desperate patients. Studying shared models like Wikipedia, Friend realized that the complexities of disease could only be understood -- and combated -- with collaboration and transparency, not by isolated scientists working in secret with proprietary data
In his quest for a solution, Friend co-founded Sage Bionetworks, an organization dedicated to creating strategies and platforms that empower researchers to share and interpret data on a colossal scale -- as well as crowdsource tests for new hypotheses.
As he wrote on CreativeCommons.org, "Our goal is ambitious. We want to take biology from a place where enclosure and privacy are the norm, where biologists see themselves as lone hunter-gatherers working to get papers written, to one where the knowledge is created specifically to fit into an open model where it can be openly queried and transformed."
Stephen Friend | Speaker | TED.com