1
00:00:00,991 --> 00:00:03,317
I get out of bed for two reasons.

2
00:00:03,341 --> 00:00:06,198
One, small-scale family farmers

3
00:00:06,222 --> 00:00:07,372
need more food.

4
00:00:07,738 --> 00:00:12,916
It's crazy that in 2019
farmers that feed us are hungry.

5
00:00:13,353 --> 00:00:18,186
And two, science needs to be
more diverse and inclusive.

6
00:00:18,679 --> 00:00:22,101
If we're going to solve
the toughest challenges on the planet,

7
00:00:22,125 --> 00:00:26,458
like food insecurity for the millions
living in extreme poverty,

8
00:00:26,482 --> 00:00:28,101
it's going to take all of us.

9
00:00:28,680 --> 00:00:31,260
I want to use the latest technology

10
00:00:31,284 --> 00:00:34,577
with the most diverse
and inclusive teams on the planet

11
00:00:34,601 --> 00:00:36,668
to help farmers have more food.

12
00:00:37,545 --> 00:00:39,426
I'm a computational biologist.

13
00:00:39,450 --> 00:00:42,854
I know - what is that
and how is it going to help end hunger?

14
00:00:42,878 --> 00:00:46,124
Basically, I like computers and biology

15
00:00:46,148 --> 00:00:47,307
and somehow,

16
00:00:47,331 --> 00:00:48,938
putting that together is a job.

17
00:00:49,529 --> 00:00:51,243
I don't have a story

18
00:00:51,267 --> 00:00:54,553
of wanting to be a biologist
from a young age.

19
00:00:54,577 --> 00:00:58,283
The truth is, I played
basketball in college.

20
00:00:58,585 --> 00:01:01,363
And part of my financial aid package

21
00:01:01,387 --> 00:01:03,728
was I needed a work-study job.

22
00:01:04,300 --> 00:01:05,840
So one random day,

23
00:01:05,864 --> 00:01:09,061
I wandered to the nearest building
to my dorm room.

24
00:01:09,085 --> 00:01:11,765
And it just so happens
it was the biology building.

25
00:01:12,347 --> 00:01:14,933
I went inside and looked at the job board.

26
00:01:15,493 --> 00:01:17,657
Yes, this is pre-the internet.

27
00:01:18,430 --> 00:01:20,437
And I saw a three-by-five card

28
00:01:20,461 --> 00:01:23,616
advertizing a job
to work in the herbarium.

29
00:01:24,601 --> 00:01:26,602
I quickly took down the number,

30
00:01:26,626 --> 00:01:28,331
because it said flexible hours,

31
00:01:28,355 --> 00:01:31,612
and I needed that to work around
my basketball schedule.

32
00:01:32,204 --> 00:01:36,791
I ran to the library
to figure out what an herbarium was.

33
00:01:36,815 --> 00:01:39,022
(Laughter)

34
00:01:39,046 --> 00:01:40,355
And it turns out,

35
00:01:40,379 --> 00:01:44,458
an herbarium is where they store
dead, dried plants.

36
00:01:45,379 --> 00:01:47,093
I was lucky to land the job.

37
00:01:47,117 --> 00:01:50,323
So my first scientific job

38
00:01:50,347 --> 00:01:55,682
was gluing dead plants onto paper
for hours on end.

39
00:01:55,706 --> 00:01:58,092
(Laughter)

40
00:01:59,008 --> 00:02:00,158
It's so glamorous.

41
00:02:00,182 --> 00:02:03,321
This is how I became
a computational biologist.

42
00:02:04,323 --> 00:02:05,506
During that time,

43
00:02:05,530 --> 00:02:08,252
genomics and computing were coming of age.

44
00:02:08,276 --> 00:02:10,680
And I went on to do my masters

45
00:02:10,704 --> 00:02:13,799
combining biology and computers.

46
00:02:13,823 --> 00:02:14,988
During that time,

47
00:02:15,012 --> 00:02:16,791
I worked at Los Alamos National lab

48
00:02:16,815 --> 00:02:19,333
in the theoretical biology
and biophysics group.

49
00:02:19,776 --> 00:02:22,101
And it was there I had my first encounter

50
00:02:22,125 --> 00:02:23,506
with the supercomputer,

51
00:02:23,530 --> 00:02:25,204
and my mind was blown.

52
00:02:25,840 --> 00:02:27,879
With the power of supercomputing,

53
00:02:27,903 --> 00:02:32,126
which is basically thousands
of connected PCs on steroids,

54
00:02:32,150 --> 00:02:34,856
we were able to uncover the complexities

55
00:02:34,880 --> 00:02:37,623
of influenza and hepatitis C.

56
00:02:38,134 --> 00:02:40,465
And it was during this time
that I saw the power

57
00:02:40,489 --> 00:02:44,609
of using computers
and biology combined, for humanity.

58
00:02:44,633 --> 00:02:47,233
And I wanted this to be my career path.

59
00:02:48,030 --> 00:02:49,807
So, since 1999,

60
00:02:49,831 --> 00:02:52,529
I've spent the majority
of my scientific career

61
00:02:52,553 --> 00:02:54,482
in very high-tech labs,

62
00:02:54,506 --> 00:02:57,239
surrounded by really expensive equipment.

63
00:02:57,712 --> 00:02:59,355
So many ask me

64
00:02:59,379 --> 00:03:03,246
how and why do I work
for farmers in Africa.

65
00:03:03,804 --> 00:03:06,106
Well, because of my computing skills,

66
00:03:06,130 --> 00:03:10,669
in 2013, a team of East African scientists

67
00:03:10,693 --> 00:03:12,439
asked me to join the team

68
00:03:12,463 --> 00:03:14,782
in the plight to save cassava.

69
00:03:15,766 --> 00:03:19,764
Cassava is a plant whose leaves and roots

70
00:03:19,788 --> 00:03:22,836
feed 800 million people globally.

71
00:03:23,639 --> 00:03:26,676
And 500 million in East Africa.

72
00:03:26,994 --> 00:03:29,001
So that's nearly a billion people

73
00:03:29,025 --> 00:03:31,993
relying on this plant
for their daily calories.

74
00:03:32,581 --> 00:03:36,426
If a small-scale family farmer
has enough cassava,

75
00:03:36,450 --> 00:03:38,594
she can feed her family

76
00:03:38,618 --> 00:03:40,561
and she can sell it at the market

77
00:03:40,585 --> 00:03:42,664
for important things like school fees,

78
00:03:42,688 --> 00:03:44,823
medical expenses and savings.

79
00:03:45,752 --> 00:03:49,283
But cassava is under attack in Africa.

80
00:03:49,665 --> 00:03:54,101
Whiteflies and viruses
are devastating cassava.

81
00:03:54,593 --> 00:03:56,799
Whiteflies are tiny insects

82
00:03:56,823 --> 00:03:59,641
that feed on the leaves
of over 600 plants.

83
00:03:59,665 --> 00:04:01,466
They are bad news.

84
00:04:01,490 --> 00:04:02,649
There are many species,

85
00:04:02,673 --> 00:04:04,942
they become pesticide resistant,

86
00:04:04,966 --> 00:04:09,220
and they transmit hundreds
of plant viruses

87
00:04:09,244 --> 00:04:11,768
that cause cassava brown streak disease

88
00:04:11,792 --> 00:04:13,592
and cassava mosaic disease.

89
00:04:14,085 --> 00:04:16,219
This completely kills the plant.

90
00:04:17,038 --> 00:04:18,855
And if there's no cassava,

91
00:04:18,879 --> 00:04:22,878
there's no food or income
for millions of people.

92
00:04:24,141 --> 00:04:26,617
It took me one trip to Tanzania

93
00:04:26,641 --> 00:04:29,379
to realize that these women
need some help.

94
00:04:29,403 --> 00:04:33,656
These amazing, strong,
small-scale family farmers,

95
00:04:33,680 --> 00:04:34,948
the majority women,

96
00:04:34,972 --> 00:04:36,239
are doing it rough.

97
00:04:36,744 --> 00:04:39,180
They don't have enough food
to feed their families,

98
00:04:39,204 --> 00:04:40,792
and it's a real crisis.

99
00:04:41,530 --> 00:04:43,029
What happens is

100
00:04:43,053 --> 00:04:46,045
they go out and plant fields of cassava
when the rains come.

101
00:04:46,069 --> 00:04:47,775
Nine months later,

102
00:04:47,799 --> 00:04:50,879
there's nothing, because of these
pests and pathogens.

103
00:04:50,903 --> 00:04:53,061
And I thought to myself,

104
00:04:53,085 --> 00:04:56,283
how in the world can farmers be hungry?

105
00:04:56,815 --> 00:04:59,135
So I decided to spend
some time on the ground

106
00:04:59,159 --> 00:05:00,839
with the farmers and the scientists

107
00:05:00,863 --> 00:05:03,466
to see if I had any skills
that could be helpful.

108
00:05:04,427 --> 00:05:07,283
The situation on the ground is shocking.

109
00:05:07,307 --> 00:05:09,504
The whiteflies have destroyed the leaves

110
00:05:09,528 --> 00:05:11,577
that are eaten for protein,

111
00:05:11,601 --> 00:05:15,183
and the viruses have destroyed the roots
that are eaten for starch.

112
00:05:15,592 --> 00:05:18,037
An entire growing season will pass,

113
00:05:18,061 --> 00:05:22,171
and the farmer will lose
an entire year of income and food

114
00:05:22,195 --> 00:05:25,393
and the family will suffer
a long hunger season.

115
00:05:25,942 --> 00:05:28,022
This is completely preventable.

116
00:05:28,046 --> 00:05:29,370
If the farmer knew

117
00:05:29,394 --> 00:05:32,458
what variety of cassava
to plant in her field,

118
00:05:32,482 --> 00:05:36,807
that was resistant to those
viruses and pathogens,

119
00:05:36,831 --> 00:05:38,736
they would have more food.

120
00:05:38,760 --> 00:05:41,595
We have all the technology we need,

121
00:05:41,619 --> 00:05:44,823
but the knowledge and the resources

122
00:05:44,847 --> 00:05:47,982
are not equally distributed
around the globe.

123
00:05:48,712 --> 00:05:51,274
So what I mean specifically is,

124
00:05:51,298 --> 00:05:53,150
the older genomic technologies

125
00:05:53,174 --> 00:05:56,037
that have been required
to uncover the complexities

126
00:05:56,061 --> 00:05:59,123
in these pests and pathogens,

127
00:05:59,147 --> 00:06:02,145
these technologies were not made
for sub-Saharan Africa.

128
00:06:03,058 --> 00:06:05,399
They cost upwards of a million dollars,

129
00:06:05,423 --> 00:06:07,311
they require constant power,

130
00:06:07,335 --> 00:06:09,135
specialized human capacity.

131
00:06:09,970 --> 00:06:12,831
These machines are few
and far between on the continent,

132
00:06:12,855 --> 00:06:17,476
which is leaving many scientists
battling on the front lines no choice

133
00:06:17,500 --> 00:06:19,499
but to send the samples overseas.

134
00:06:19,523 --> 00:06:21,483
And when you send the samples overseas,

135
00:06:21,507 --> 00:06:24,133
samples degrade, it costs a lot of money,

136
00:06:24,157 --> 00:06:27,324
and trying to get the data back
over weak internet

137
00:06:27,348 --> 00:06:28,748
is nearly impossible.

138
00:06:29,142 --> 00:06:31,362
So sometimes it can take six months

139
00:06:31,386 --> 00:06:33,441
to get the results back to the farmer.

140
00:06:33,465 --> 00:06:35,219
And by then it's too late.

141
00:06:35,243 --> 00:06:36,830
The crop is already gone,

142
00:06:36,854 --> 00:06:40,020
which results in further poverty
and more hunger.

143
00:06:41,306 --> 00:06:43,464
We knew we could fix this.

144
00:06:43,989 --> 00:06:45,393
In 2017,

145
00:06:45,417 --> 00:06:50,203
we had heard of this hand-held,
portable DNA sequencer

146
00:06:50,227 --> 00:06:52,736
called an Oxford nanopore MinION.

147
00:06:52,760 --> 00:06:56,913
This was being used
in West Africa to fight Ebola.

148
00:06:56,937 --> 00:06:58,434
So we thought,

149
00:06:58,458 --> 00:07:01,744
why can't we use this
in East Africa to help farmers?

150
00:07:01,768 --> 00:07:06,101
So, what we did was we set out to do that.

151
00:07:06,609 --> 00:07:09,307
At the time, the technology was very new,

152
00:07:09,331 --> 00:07:12,283
and many doubted we could
replicate this on the farm.

153
00:07:12,879 --> 00:07:14,196
When we set out to do this,

154
00:07:14,220 --> 00:07:18,101
one of our "collaborators" in the UK

155
00:07:18,125 --> 00:07:21,752
told us that we would never
get that to work in East Africa,

156
00:07:21,776 --> 00:07:23,242
let alone on the farm.

157
00:07:23,863 --> 00:07:25,632
So we accepted the challenge.

158
00:07:25,934 --> 00:07:29,791
This person even went so far as to bet us

159
00:07:29,815 --> 00:07:32,387
two of the best bottles of champagne

160
00:07:32,411 --> 00:07:35,369
that we would never get that to work.

161
00:07:36,871 --> 00:07:38,450
Two words:

162
00:07:38,474 --> 00:07:39,625
pay up.

163
00:07:39,649 --> 00:07:41,823
(Laughter)

164
00:07:41,847 --> 00:07:45,999
(Applause)

165
00:07:46,023 --> 00:07:48,936
Pay up, because we did it.

166
00:07:48,960 --> 00:07:52,245
We took the entire high-tech molecular lab

167
00:07:52,269 --> 00:07:55,918
to the farmers of Tanzania,
Kenya and Uganda,

168
00:07:55,942 --> 00:07:57,974
and we called it Tree Lab.

169
00:07:58,942 --> 00:08:00,133
So what did we do?

170
00:08:00,157 --> 00:08:02,736
Well, first of all,
we gave ourselves a team name,

171
00:08:02,760 --> 00:08:04,934
it's called the Cassava Virus
Action Project,

172
00:08:04,958 --> 00:08:06,315
we made a website,

173
00:08:06,339 --> 00:08:09,950
we gathered support from the genomics
and computing communities,

174
00:08:09,974 --> 00:08:11,855
and away we went to the farmers.

175
00:08:12,411 --> 00:08:15,220
Everything that we need for our Tree Lab

176
00:08:15,244 --> 00:08:17,653
is being carried by the team here.

177
00:08:17,677 --> 00:08:21,724
All of the molecular and computational
requirements needed

178
00:08:21,748 --> 00:08:25,049
to diagnose sick plants is there.

179
00:08:25,431 --> 00:08:28,259
And it's actually all
on this stage here as well.

180
00:08:29,161 --> 00:08:32,748
We figured if we could get the data
closer to the problem,

181
00:08:32,772 --> 00:08:34,390
and closer to the farmer,

182
00:08:34,414 --> 00:08:37,770
the quicker we could tell her
what was wrong with her plant.

183
00:08:38,169 --> 00:08:39,942
And not only tell her what was wrong,

184
00:08:39,966 --> 00:08:41,457
give her the solution.

185
00:08:41,482 --> 00:08:42,807
And the solution is,

186
00:08:42,831 --> 00:08:45,454
burn the field and plant varieties

187
00:08:45,478 --> 00:08:48,982
that are resistant to the pests
and pathogens she has in her field.

188
00:08:49,942 --> 00:08:54,146
So the first thing that we did
was we had to do a DNA extraction.

189
00:08:54,170 --> 00:08:56,709
And we used this machine here.

190
00:08:57,050 --> 00:08:59,940
It's called a PDQEX.

191
00:09:00,273 --> 00:09:04,164
Which stands for
Pretty Damn Quick Extraction.

192
00:09:04,188 --> 00:09:06,236
(Laughter)

193
00:09:06,260 --> 00:09:07,410
I know.

194
00:09:07,768 --> 00:09:10,262
My friend Joe is really cool.

195
00:09:11,394 --> 00:09:13,069
One of the biggest challenges

196
00:09:13,093 --> 00:09:16,347
in doing a DNA extraction
is it usually requires

197
00:09:16,371 --> 00:09:18,093
very expensive equipment,

198
00:09:18,117 --> 00:09:19,521
and takes hours.

199
00:09:19,545 --> 00:09:21,037
But with this machine

200
00:09:21,061 --> 00:09:23,815
we've been able to do it in 20 minutes

201
00:09:23,839 --> 00:09:25,085
at a fraction of the cost.

202
00:09:25,109 --> 00:09:27,997
And this runs off of a motorcycle battery.

203
00:09:29,164 --> 00:09:31,773
From there, we take the DNA extraction

204
00:09:31,797 --> 00:09:34,307
and prepare it into a library,

205
00:09:34,331 --> 00:09:36,110
getting it ready to load on

206
00:09:36,134 --> 00:09:40,426
to this portable, hand-held
genomic sequencer,

207
00:09:40,450 --> 00:09:41,601
which is here,

208
00:09:41,625 --> 00:09:45,363
and then we plug this
into a mini supercomputer,

209
00:09:45,387 --> 00:09:47,209
which is called a MinIT.

210
00:09:47,728 --> 00:09:51,830
And both of these things are plugged in
to a portable battery pack.

211
00:09:52,569 --> 00:09:54,442
So we were able to eliminate

212
00:09:54,466 --> 00:09:56,871
the requirements
of main power and internet,

213
00:09:56,895 --> 00:10:00,823
which are two very limiting factors
on a small-scale family farm.

214
00:10:01,807 --> 00:10:04,678
Analyzing the data quickly
can also be a problem.

215
00:10:05,033 --> 00:10:08,939
But this is where me being
a computational biologist came in handy.

216
00:10:09,382 --> 00:10:11,612
All that gluing of dead plants,

217
00:10:11,636 --> 00:10:13,196
and all that measuring,

218
00:10:13,220 --> 00:10:15,212
and all that computing,

219
00:10:15,236 --> 00:10:17,022
finally came in handy

220
00:10:17,046 --> 00:10:19,387
in a real-world, real-time way.

221
00:10:19,411 --> 00:10:22,464
I was able to make customized data bases

222
00:10:22,488 --> 00:10:27,083
and we were able to give the farmers
results in three hours

223
00:10:27,107 --> 00:10:28,971
versus six months.

224
00:10:30,108 --> 00:10:37,076
(Applause)

225
00:10:38,085 --> 00:10:40,719
The farmers were overjoyed.

226
00:10:41,799 --> 00:10:44,595
So how do we know
that we're having impact?

227
00:10:44,619 --> 00:10:46,619
Nine moths after our Tree Lab,

228
00:10:46,643 --> 00:10:49,873
Asha went from having
zero tons per hectare

229
00:10:49,897 --> 00:10:51,905
to 40 tons per hectare.

230
00:10:51,929 --> 00:10:53,728
She had enough to feed her family

231
00:10:53,752 --> 00:10:56,442
and she was selling it at the market

232
00:10:56,466 --> 00:10:59,201
and she's now building a house
for her family.

233
00:11:00,212 --> 00:11:01,371
Yeah, so cool.

234
00:11:01,395 --> 00:11:05,649
(Applause)

235
00:11:05,673 --> 00:11:07,539
So how do we scale Tree Lab?

236
00:11:07,940 --> 00:11:09,320
The thing is,

237
00:11:09,344 --> 00:11:11,447
farmers are scaled already in Africa.

238
00:11:11,471 --> 00:11:13,360
These women work in farmer groups,

239
00:11:13,384 --> 00:11:17,510
so helping Asha actually helped
3,000 people in her village,

240
00:11:17,534 --> 00:11:21,186
because she shared the results
and also the solution.

241
00:11:21,673 --> 00:11:25,864
I remember every single
farmer I've ever met.

242
00:11:26,665 --> 00:11:30,228
Their pain and their joy

243
00:11:30,252 --> 00:11:32,052
is engraved in my memories.

244
00:11:32,958 --> 00:11:34,823
Our science is for them.

245
00:11:35,711 --> 00:11:38,268
Tree Lab is our best attempt

246
00:11:38,292 --> 00:11:40,758
to help them become more food-secure.

247
00:11:41,180 --> 00:11:42,966
I never dreamt

248
00:11:42,990 --> 00:11:45,934
that the best science
I would ever do in my life

249
00:11:45,958 --> 00:11:49,457
would be on that blanket in East Africa,

250
00:11:49,481 --> 00:11:51,847
with the highest tech genomic gadgets.

251
00:11:52,312 --> 00:11:54,764
But our team did dream

252
00:11:54,788 --> 00:11:59,058
that we could give farmers answers
in three hours versus six months,

253
00:11:59,082 --> 00:12:00,518
and then we did it.

254
00:12:00,542 --> 00:12:04,650
Because that's the power
of diversity and inclusion inn science.

255
00:12:05,156 --> 00:12:06,307
Thank you.

256
00:12:06,331 --> 00:12:09,482
(Applause)

257
00:12:09,506 --> 00:12:13,589
(Cheers)