1
00:00:00,980 --> 00:00:01,910
- Welcome everyone,

2
00:00:01,910 --> 00:00:05,590
to the 43rd annual
Darwin festival committee

3
00:00:05,590 --> 00:00:08,710
with tonight's speaker
being Dr. Sarah Tishkoff.

4
00:00:08,710 --> 00:00:10,940
I will hand the microphone
over to my colleague,

5
00:00:10,940 --> 00:00:13,483
Dr. Jason Brown, to make the introduction.

6
00:00:15,290 --> 00:00:18,420
- Welcome to Salem State, students,

7
00:00:18,420 --> 00:00:20,990
faculty, staff, and alumni.

8
00:00:20,990 --> 00:00:25,990
And to our visitors and
our speaker, Dr. Tishkoff.

9
00:00:26,280 --> 00:00:28,520
Sarah Tishkoff has a Bachelor of Science

10
00:00:28,520 --> 00:00:30,580
in Anthropology and Genetics

11
00:00:30,580 --> 00:00:33,530
from the University of
California, Berkeley

12
00:00:33,530 --> 00:00:35,440
and a Masters in Human Genetics,

13
00:00:35,440 --> 00:00:39,060
and a PhD in Genetics
from Yale University.

14
00:00:39,060 --> 00:00:42,330
She's the David and Lyn
Silfen University Professor

15
00:00:42,330 --> 00:00:45,840
in genetics and biology at the
University of Pennsylvania,

16
00:00:45,840 --> 00:00:47,360
and is director of

17
00:00:47,360 --> 00:00:51,250
the Penn Center for Global
Genomics & Health Equity.

18
00:00:51,250 --> 00:00:54,250
She's a co-author of over 130 publications

19
00:00:54,250 --> 00:00:58,470
with a primary focus on
African genetic diversity

20
00:00:58,470 --> 00:01:00,510
and human health.

21
00:01:00,510 --> 00:01:01,840
Dr. Tishkoff is a member

22
00:01:01,840 --> 00:01:03,990
of the National Academy of Sciences,

23
00:01:03,990 --> 00:01:06,830
where she's on the board of global health,

24
00:01:06,830 --> 00:01:07,770
and she received

25
00:01:07,770 --> 00:01:11,080
a National Institutes
of Health Pioneer Award

26
00:01:11,080 --> 00:01:14,240
along with multiple
other prestigious awards.

27
00:01:14,240 --> 00:01:19,240
And she's on the editorial
board of several top journals.

28
00:01:19,320 --> 00:01:21,790
So my students and I
are particularly excited

29
00:01:21,790 --> 00:01:24,490
that Dr. Tishkoff was able
to join us this semester

30
00:01:24,490 --> 00:01:26,720
during the inaugural run

31
00:01:26,720 --> 00:01:30,750
of our new genetics of
human disease course.

32
00:01:30,750 --> 00:01:34,720
So with no further delay Dr. Tishkoff.

33
00:01:34,720 --> 00:01:37,740
- Well, thank you so much for
that excellent introduction.

34
00:01:37,740 --> 00:01:39,393
Lemme just get my slides up.

35
00:01:41,240 --> 00:01:43,740
Let me put it on presenter mode.

36
00:01:43,740 --> 00:01:48,510
I have to say that it is a
particular honor to be giving

37
00:01:48,510 --> 00:01:51,690
this talk today in
honor of Charles Darwin,

38
00:01:51,690 --> 00:01:55,210
whose birthday is just a couple days away.

39
00:01:55,210 --> 00:01:58,330
And I thought I would start
by talking about what I think

40
00:01:58,330 --> 00:02:01,753
are some of the key challenges
in human genomics research.

41
00:02:03,140 --> 00:02:05,810
And that includes that we
need to do much better job

42
00:02:05,810 --> 00:02:09,210
characterizing genomic
and phenotypic variation,

43
00:02:09,210 --> 00:02:10,510
ethnically diverse humans.

44
00:02:10,510 --> 00:02:12,840
What do I mean by phenotypic variation?

45
00:02:12,840 --> 00:02:16,783
Just that could be normal,
trade variation or disease risk.

46
00:02:18,330 --> 00:02:20,720
We wanna understand the
evolutionary processes

47
00:02:20,720 --> 00:02:23,690
that generate and maintain that variation,

48
00:02:23,690 --> 00:02:27,220
and to understand gene-gene, gene-protein,

49
00:02:27,220 --> 00:02:29,900
and gene-environment interactions
and how they contribute

50
00:02:29,900 --> 00:02:32,483
both to normal variation and disease risk.

51
00:02:33,500 --> 00:02:36,363
The focus of our research is on Africa,

52
00:02:37,290 --> 00:02:40,020
and there are number of reasons for that.

53
00:02:40,020 --> 00:02:41,360
One of which is the fact

54
00:02:41,360 --> 00:02:44,770
that all modern humans
originated in Africa.

55
00:02:44,770 --> 00:02:47,000
These red dots represent the locations

56
00:02:47,000 --> 00:02:50,110
of fossils of anatomically modern humans.

57
00:02:50,110 --> 00:02:53,457
The oldest of which is dated
to about 300,000 years ago.

58
00:02:55,800 --> 00:02:58,360
We also have some archeological evidence

59
00:02:58,360 --> 00:03:03,030
for early modern humans
from early modern humans.

60
00:03:03,030 --> 00:03:06,400
So for example, this is
from a cave in South Africa,

61
00:03:06,400 --> 00:03:08,250
and it was one of the earliest examples

62
00:03:08,250 --> 00:03:10,030
where you could see that they,

63
00:03:10,030 --> 00:03:13,290
somebody has etched into this stone.

64
00:03:13,290 --> 00:03:16,163
And that was around 70,000 years ago.

65
00:03:17,730 --> 00:03:19,120
And then this was pretty neat.

66
00:03:19,120 --> 00:03:20,260
They found what they call

67
00:03:20,260 --> 00:03:22,550
an artist toolkit in the same cave,

68
00:03:22,550 --> 00:03:24,550
but it was actually dated to a bit earlier

69
00:03:24,550 --> 00:03:27,230
around 100,000 years ago,

70
00:03:27,230 --> 00:03:29,530
and they found some pigments in there,

71
00:03:29,530 --> 00:03:32,440
which indicated that they
were using the shell to mix up

72
00:03:32,440 --> 00:03:34,730
pigments either for cave painting,

73
00:03:34,730 --> 00:03:36,753
or for painting their own bodies.

74
00:03:39,110 --> 00:03:42,470
So after this origin in Africa,

75
00:03:42,470 --> 00:03:45,420
somewhere around 50 to 80,000 years ago,

76
00:03:45,420 --> 00:03:49,090
relatively small numbers of
people migrated out of Africa

77
00:03:49,090 --> 00:03:50,393
and across the globe.

78
00:03:51,240 --> 00:03:54,000
And when they left Africa,

79
00:03:54,000 --> 00:03:56,110
they ran into archaic populations

80
00:03:56,110 --> 00:03:59,030
like Neanderthals and Denisovans,

81
00:03:59,030 --> 00:04:02,900
And we now know that they
interbred to some small extent,

82
00:04:02,900 --> 00:04:06,780
such that somewhere between
about two to 4% of the genomes

83
00:04:06,780 --> 00:04:09,380
of people outside of
Africa are actually derived

84
00:04:09,380 --> 00:04:11,573
from those archaic populations.

85
00:04:13,580 --> 00:04:17,313
So how much do we differ at
the level of our genomes?

86
00:04:18,640 --> 00:04:22,803
So identical twins should
have no differences in theory.

87
00:04:24,360 --> 00:04:28,330
If we compared the genome,
all the genetic material

88
00:04:28,330 --> 00:04:32,530
in the nuclear part of
your cell in the nucleus,

89
00:04:32,530 --> 00:04:36,405
then we would differed
about one out of 1000 sites,

90
00:04:36,405 --> 00:04:39,793
approximately, nucleotide sites.

91
00:04:40,750 --> 00:04:44,330
If we compared our genome, switch it,

92
00:04:44,330 --> 00:04:46,793
we differed about one out of 100 sites.

93
00:04:47,810 --> 00:04:50,863
And if we compare it to a
mouse, it's about one out of 30.

94
00:04:51,810 --> 00:04:53,250
And if we compare it to broccoli,

95
00:04:53,250 --> 00:04:55,410
it's about two out of three.

96
00:04:55,410 --> 00:04:59,430
So given that there are
about 3,000,000,000 DNA bases

97
00:04:59,430 --> 00:05:03,520
in the genome, that's
about 3,000,000 differences

98
00:05:03,520 --> 00:05:06,703
between genomes, between
different individuals.

99
00:05:07,960 --> 00:05:10,640
Now, I like to use this slide.

100
00:05:10,640 --> 00:05:11,473
It's sort of funny.

101
00:05:11,473 --> 00:05:14,160
I say it's a representation
of ethnic diversity

102
00:05:14,160 --> 00:05:17,450
amongst beautiful people, but
even amongst the rest of us,

103
00:05:17,450 --> 00:05:21,560
there's less than 0.1
divergence or differences

104
00:05:21,560 --> 00:05:24,440
at the genome's wide level.

105
00:05:24,440 --> 00:05:26,087
And that's reflecting this very recent

106
00:05:26,087 --> 00:05:28,093
common ancestry in Africa.

107
00:05:29,630 --> 00:05:32,510
Now there are other types
of variation in the genome.

108
00:05:32,510 --> 00:05:34,300
We can call them structural variants.

109
00:05:34,300 --> 00:05:36,530
So there could be chunks of the chromosome

110
00:05:36,530 --> 00:05:39,830
that are inserted, or
deleted, or moved around.

111
00:05:39,830 --> 00:05:41,690
Those are not super well characterized.

112
00:05:41,690 --> 00:05:43,170
They actually could be contributing

113
00:05:43,170 --> 00:05:45,330
to a lot of variation between individuals,

114
00:05:45,330 --> 00:05:49,000
but we still have a long way
to ago to characterize those.

115
00:05:49,000 --> 00:05:53,410
But every study for the past
50 years has shown consistently

116
00:05:53,410 --> 00:05:57,210
that the majority of
genetic variation in humans

117
00:05:57,210 --> 00:06:00,260
is within populations, 85%,

118
00:06:00,260 --> 00:06:03,730
relative to between
populations around 15%.

119
00:06:03,730 --> 00:06:07,033
Again, reflecting this
recent African origin.

120
00:06:08,550 --> 00:06:11,810
So why should we study
African genetic diversity?

121
00:06:11,810 --> 00:06:14,280
Well, one is to reconstruct
human demographic

122
00:06:14,280 --> 00:06:15,913
and evolutionary history,

123
00:06:17,160 --> 00:06:20,943
to study the African diaspora
and African American ancestry,

124
00:06:22,070 --> 00:06:25,350
to study the genetic
basis of susceptibility

125
00:06:25,350 --> 00:06:29,510
to both communicable and
non-communicable diseases.

126
00:06:29,510 --> 00:06:32,830
For example, malaria and TB
are very common in Africa,

127
00:06:32,830 --> 00:06:36,340
but increasingly diseases
like type two diabetes

128
00:06:36,340 --> 00:06:38,130
and hypertension are also on the rise,

129
00:06:38,130 --> 00:06:39,783
particularly in urban areas.

130
00:06:41,090 --> 00:06:44,480
And to understand things like
differences in drug response

131
00:06:44,480 --> 00:06:46,973
that could contribute
to precision medicine.

132
00:06:49,440 --> 00:06:54,380
And yet there's a major
bias towards Europeans

133
00:06:54,380 --> 00:06:56,400
in genetics dataset.

134
00:06:56,400 --> 00:06:58,170
So this is actually from a dataset

135
00:06:59,110 --> 00:07:02,880
where they look for genetic
associations with disease.

136
00:07:02,880 --> 00:07:04,350
And from that dataset,

137
00:07:04,350 --> 00:07:07,470
we could see that about 80%
of the individuals included

138
00:07:07,470 --> 00:07:09,400
are of European ancestry.

139
00:07:09,400 --> 00:07:12,830
About 10% are of Asian ancestry.

140
00:07:12,830 --> 00:07:15,450
2% are of African ancestry,

141
00:07:15,450 --> 00:07:19,840
about 1% Hispanic ancestry and
less than 1% everybody else.

142
00:07:19,840 --> 00:07:23,390
And that's really a problem
both in terms of learning about

143
00:07:23,390 --> 00:07:25,040
our human evolutionary history,

144
00:07:25,040 --> 00:07:29,790
but also in terms of more
equitable healthcare,

145
00:07:29,790 --> 00:07:31,833
I would say, and health benefits.

146
00:07:33,620 --> 00:07:36,080
So I wanna tell you a bit about Africa.

147
00:07:36,080 --> 00:07:38,430
There are over 2000
ethnic groups in Africa,

148
00:07:38,430 --> 00:07:40,440
speaking languages that
have been classified

149
00:07:40,440 --> 00:07:42,970
into four major language families.

150
00:07:42,970 --> 00:07:46,710
In blue, I'm showing the
distribution of languages

151
00:07:46,710 --> 00:07:48,870
that have been classified as Afro-Asiatic.

152
00:07:48,870 --> 00:07:52,190
These are mainly in
Northern and Eastern Africa.

153
00:07:52,190 --> 00:07:54,490
In red is the distribution of populations

154
00:07:54,490 --> 00:07:57,620
that speak Nilo-Saharan languages,

155
00:07:57,620 --> 00:08:00,450
mainly in Central and Eastern Africa.

156
00:08:00,450 --> 00:08:02,580
These groups tend to be pastoralists,

157
00:08:02,580 --> 00:08:05,130
so they raise cattle and they drink milk.

158
00:08:05,130 --> 00:08:08,290
An example you might have
heard of are the Maasai.

159
00:08:08,290 --> 00:08:11,400
The most widespread language
family is Niger-Kordofanian,

160
00:08:11,400 --> 00:08:16,080
or Niger-Congo that
originated in Western Africa,

161
00:08:16,080 --> 00:08:20,700
and the largest sub-family,
sub-language family,

162
00:08:20,700 --> 00:08:21,980
I guess you could say,

163
00:08:21,980 --> 00:08:24,010
are the Bantu languages that are thought

164
00:08:24,010 --> 00:08:26,220
to have originated around along the border

165
00:08:26,220 --> 00:08:28,680
of Cameroon in Nigeria.

166
00:08:28,680 --> 00:08:32,110
So these populations had
developed iron tool technology,

167
00:08:32,110 --> 00:08:35,490
and something called slash
and burn agriculture.

168
00:08:35,490 --> 00:08:38,900
So they were very successful
at cutting down the forest,

169
00:08:38,900 --> 00:08:42,850
growing crops, sustaining
large population sizes.

170
00:08:42,850 --> 00:08:45,750
And somewhere in the
past 4,000 years or so,

171
00:08:45,750 --> 00:08:48,510
they migrated to the east
and then to the south,

172
00:08:48,510 --> 00:08:50,443
and to the west and to the south.

173
00:08:51,680 --> 00:08:54,510
Then they interbred
with local populations,

174
00:08:54,510 --> 00:08:57,950
and have really shaped the
genomic landscape in Africa.

175
00:08:57,950 --> 00:09:00,740
And then in green, we see populations

176
00:09:00,740 --> 00:09:02,010
that have been labeled as Khoisan.

177
00:09:02,010 --> 00:09:04,340
These are groups that
speak with the click.

178
00:09:04,340 --> 00:09:07,660
They would include the San
population from Southern Africa,

179
00:09:07,660 --> 00:09:10,480
who until recently have been living

180
00:09:10,480 --> 00:09:12,913
a traditionally hunting
and gathering lifestyle.

181
00:09:13,850 --> 00:09:16,430
And then in Tanzania,
there are two groups,

182
00:09:16,430 --> 00:09:18,860
the Hadza and the Sandawe
who speak with clicks

183
00:09:18,860 --> 00:09:23,130
and also practice a hunting
and gathering lifestyle,

184
00:09:23,130 --> 00:09:24,843
at least until recently.

185
00:09:27,310 --> 00:09:31,150
So all of this research has
been done in partnership

186
00:09:31,150 --> 00:09:34,590
with a number of faculty in
different countries in Africa.

187
00:09:34,590 --> 00:09:36,350
And I'm gonna show you a little bit

188
00:09:36,350 --> 00:09:38,883
of what the field work
is like that we've done.

189
00:09:40,610 --> 00:09:44,970
So we are mainly studying
genetic diversity from groups

190
00:09:44,970 --> 00:09:47,920
that are minority populations in Africa.

191
00:09:47,920 --> 00:09:50,730
So they tend to live in more rural areas.

192
00:09:50,730 --> 00:09:52,590
It can be challenging
to get to these groups,

193
00:09:52,590 --> 00:09:54,690
requires bringing a four
wheel drive vehicle,

194
00:09:54,690 --> 00:09:57,083
we have to bring all of
our supplies with us.

195
00:09:58,070 --> 00:09:59,690
Here you can see,

196
00:09:59,690 --> 00:10:00,990
this is sort of how we're doing

197
00:10:00,990 --> 00:10:04,290
the biometric measurements
that we look at.

198
00:10:04,290 --> 00:10:06,900
This individual in the upper
right and the lower left,

199
00:10:06,900 --> 00:10:10,614
these are members of the Hadza
tribe who speak with a click

200
00:10:10,614 --> 00:10:12,010
and this woman in the upper left,

201
00:10:12,010 --> 00:10:16,743
she speaks Nilo-Saharan language
and practices pastoralism.

202
00:10:18,850 --> 00:10:20,430
This is from Botswana.

203
00:10:20,430 --> 00:10:23,283
You can just see some
of the ethnic diversity.

204
00:10:25,400 --> 00:10:28,030
And this is a more recent
expedition to Cameroon.

205
00:10:28,030 --> 00:10:29,060
Some of our colleagues from

206
00:10:29,060 --> 00:10:31,030
the University of Yaoundé Medical School

207
00:10:31,030 --> 00:10:34,243
and people in my lab who have
contributed to this research.

208
00:10:36,060 --> 00:10:38,610
So this can be very challenging.

209
00:10:38,610 --> 00:10:41,060
This is just one example
of challenge we face,

210
00:10:41,060 --> 00:10:42,450
getting up these roads

211
00:10:42,450 --> 00:10:44,853
when they're just completely muddy.

212
00:10:47,130 --> 00:10:49,670
I remember this bridge in
Cameroon and it would terrify me.

213
00:10:49,670 --> 00:10:51,390
I just was like so scared

214
00:10:51,390 --> 00:10:54,090
that we were gonna go
crash into the river.

215
00:10:54,090 --> 00:10:55,910
And then the other thing in Cameroon

216
00:10:55,910 --> 00:10:58,060
is that they have these
giant logging trucks,

217
00:10:58,060 --> 00:10:59,140
and when they're in front of you,

218
00:10:59,140 --> 00:11:01,870
they're going like 5 mph
and you wanna pass them,

219
00:11:01,870 --> 00:11:05,017
but on the dirt roads, they're
kicking up all this dirt

220
00:11:05,017 --> 00:11:06,230
and you can't see.

221
00:11:06,230 --> 00:11:07,600
And then when they're behind you,

222
00:11:07,600 --> 00:11:10,403
they're going like 100
mph and it's really scary.

223
00:11:12,280 --> 00:11:16,620
When we do this research,
we are particularly careful

224
00:11:16,620 --> 00:11:18,720
to do this in an ethical manner.

225
00:11:18,720 --> 00:11:22,010
So that means that we go
through ethical review

226
00:11:22,010 --> 00:11:25,150
and we get IRB approval at
our university in the US,

227
00:11:25,150 --> 00:11:28,670
but we also have to go through
many rounds of ethical review

228
00:11:28,670 --> 00:11:32,070
in each country until we
get a research permit.

229
00:11:32,070 --> 00:11:36,080
We also spend a lot of time
having community discussions,

230
00:11:36,080 --> 00:11:38,390
we explain the project in layman's terms,

231
00:11:38,390 --> 00:11:40,300
we translate into the local language,

232
00:11:40,300 --> 00:11:43,180
we talk about the risks
and the benefits if any,

233
00:11:43,180 --> 00:11:45,090
it's only after getting community consent

234
00:11:45,090 --> 00:11:46,870
and ultimately individual consent

235
00:11:46,870 --> 00:11:48,623
that we proceed with the research.

236
00:11:50,080 --> 00:11:51,930
We also think that it's really important

237
00:11:51,930 --> 00:11:54,720
to return results to participants,

238
00:11:54,720 --> 00:11:56,763
and this is a way of benefit sharing.

239
00:11:58,440 --> 00:12:01,020
Training and capacity building
is also really important

240
00:12:01,020 --> 00:12:01,853
and I've had the honor

241
00:12:01,853 --> 00:12:03,610
of training a number of grad students

242
00:12:03,610 --> 00:12:05,653
and post-docs from Africa.

243
00:12:07,220 --> 00:12:11,290
So in these populations,
we're mainly obtaining blood,

244
00:12:11,290 --> 00:12:15,810
and from blood you can get
DNA and RNA and plasma.

245
00:12:15,810 --> 00:12:17,550
We also look at fecal samples,

246
00:12:17,550 --> 00:12:19,620
so look at the gut microbiome,

247
00:12:19,620 --> 00:12:22,430
we get detailed ethnographic information,

248
00:12:22,430 --> 00:12:23,670
information about diet,

249
00:12:23,670 --> 00:12:25,963
any health information that they may have.

250
00:12:27,560 --> 00:12:28,960
And then we have to process

251
00:12:28,960 --> 00:12:31,470
these samples without electricity.

252
00:12:31,470 --> 00:12:34,600
And so what we typically do
now is just bring a generator,

253
00:12:34,600 --> 00:12:38,070
as shown in the lower
left, out to the bush,

254
00:12:38,070 --> 00:12:40,543
and we can just set up our lab anywhere.

255
00:12:42,210 --> 00:12:45,580
So we also are measuring
phenotypic variation

256
00:12:45,580 --> 00:12:48,600
in whatever we can do
in a noninvasive manner,

257
00:12:48,600 --> 00:12:50,780
in a very rural setting.

258
00:12:50,780 --> 00:12:53,700
So this includes very detailed
anthropometric phenotypes

259
00:12:53,700 --> 00:12:56,290
like height and weight and limb length,

260
00:12:56,290 --> 00:12:59,720
and we have looked at cardiovascular, lung

261
00:12:59,720 --> 00:13:02,563
and blood phenotypes, like
blood pressure for example.

262
00:13:04,090 --> 00:13:06,860
We look at metabolic
function, lactose tolerance,

263
00:13:06,860 --> 00:13:11,823
glucose tolerance and infectious
disease status when we can.

264
00:13:13,491 --> 00:13:15,210
And then we are using different types

265
00:13:15,210 --> 00:13:17,890
of Omex technologies in my lab.

266
00:13:17,890 --> 00:13:21,150
We integrate these together
and also look at the impact

267
00:13:21,150 --> 00:13:24,690
of diet and other environmental factors

268
00:13:24,690 --> 00:13:26,010
and how those are influencing

269
00:13:26,010 --> 00:13:29,123
both normal variable
traits and disease risk.

270
00:13:30,410 --> 00:13:32,790
So I wanna tell you first
about a paper we published

271
00:13:32,790 --> 00:13:36,180
a number of years ago that
looked at highly variable,

272
00:13:36,180 --> 00:13:39,420
they're called Microsatellites
short tandem repeats.

273
00:13:39,420 --> 00:13:43,570
There are little repeats
of two to six nucleotides,

274
00:13:43,570 --> 00:13:46,110
an example would be CA, CA, CA, CA,

275
00:13:46,110 --> 00:13:49,530
repeated multiple times,
and they're highly variable.

276
00:13:49,530 --> 00:13:52,290
These are what are used,
for example, in forensics,

277
00:13:52,290 --> 00:13:53,460
because they're so variable,

278
00:13:53,460 --> 00:13:55,700
they're really good genetic marker.

279
00:13:55,700 --> 00:13:59,240
And we looked at this
in over 2,500 African

280
00:13:59,240 --> 00:14:02,430
from 121 ethnic groups shown here,

281
00:14:02,430 --> 00:14:06,190
and in '98, African Americans
from four regions in the US,

282
00:14:06,190 --> 00:14:08,050
and then a large comparative data set

283
00:14:08,050 --> 00:14:09,863
from non-African individuals.

284
00:14:11,210 --> 00:14:14,120
So I first wanna show you
this slide is representing

285
00:14:14,120 --> 00:14:16,610
a measure of genetic diversity,

286
00:14:16,610 --> 00:14:18,870
and it's color coded
by geographic regions,

287
00:14:18,870 --> 00:14:21,280
so we could see that the
most genetic diversity

288
00:14:21,280 --> 00:14:23,400
is in Africa.

289
00:14:23,400 --> 00:14:26,580
We see decreasing diversity
as we go west to east,

290
00:14:26,580 --> 00:14:31,580
across Eurasia, into East
Asia, Oceania and the Americas.

291
00:14:31,750 --> 00:14:34,340
And that's reflecting that
migration I told you about.

292
00:14:34,340 --> 00:14:37,580
So when people, when ancestor...

293
00:14:37,580 --> 00:14:40,580
When people left to Africa
and there was a bottleneck,

294
00:14:40,580 --> 00:14:42,560
you see a loss of diversity,

295
00:14:42,560 --> 00:14:45,510
and then as they go into different areas

296
00:14:45,510 --> 00:14:47,580
like into Oceania or into the Americas,

297
00:14:47,580 --> 00:14:50,073
again you see a series of founding events.

298
00:14:52,180 --> 00:14:54,180
This is a phylogenetic tree.

299
00:14:54,180 --> 00:14:55,500
You can't see any details.

300
00:14:55,500 --> 00:14:58,010
I'm just gonna point
out the general trends.

301
00:14:58,010 --> 00:15:02,350
And it's constructed by
estimating the genetic distances

302
00:15:02,350 --> 00:15:05,300
between pairs of populations.

303
00:15:05,300 --> 00:15:09,000
So at the ends of each of
these branches is a population,

304
00:15:09,000 --> 00:15:11,160
the branches are color coded according

305
00:15:11,160 --> 00:15:12,790
to the language spoken in Africa.

306
00:15:12,790 --> 00:15:14,570
These are the African populations,

307
00:15:14,570 --> 00:15:18,500
and a pair of the non-African populations.

308
00:15:18,500 --> 00:15:21,110
And we could see that at
the root of this tree,

309
00:15:21,110 --> 00:15:24,090
the most basal lineages

310
00:15:24,090 --> 00:15:26,200
are those from the San hunter-gatherers,

311
00:15:26,200 --> 00:15:28,210
from Southern Africa,

312
00:15:28,210 --> 00:15:30,270
followed by so-called Pygmies,

313
00:15:30,270 --> 00:15:32,540
Central African hunter-gatherers,

314
00:15:32,540 --> 00:15:34,730
and then we see that people
largely are clustering

315
00:15:34,730 --> 00:15:36,600
by geographic region.

316
00:15:36,600 --> 00:15:41,600
Western and central Africa,
and then here Eastern Africa.

317
00:15:42,460 --> 00:15:45,010
And they generally
cluster based on language,

318
00:15:45,010 --> 00:15:47,260
but with some exceptions.

319
00:15:47,260 --> 00:15:51,560
And then we could see that
outside of Africa here is India.

320
00:15:51,560 --> 00:15:53,540
This is actually blocked in my screen.

321
00:15:53,540 --> 00:15:57,890
I think it's Central Asia,
Europe, East Asia, Americas,

322
00:15:57,890 --> 00:16:01,070
Oceania, and then we have North Africa.

323
00:16:01,070 --> 00:16:03,670
Again, that's consistent
with that out of Africa,

324
00:16:03,670 --> 00:16:05,263
model of human origins.

325
00:16:06,500 --> 00:16:08,950
We then used a computational approach

326
00:16:08,950 --> 00:16:11,680
to infer genetic ancestry,

327
00:16:11,680 --> 00:16:13,870
which is represented by
the different colors.

328
00:16:13,870 --> 00:16:16,830
So the way to look at this is that,

329
00:16:16,830 --> 00:16:19,540
you could see that this is
made up of a bunch of lines

330
00:16:19,540 --> 00:16:22,150
and each line represents a person.

331
00:16:22,150 --> 00:16:25,300
And for each person we've
inferred their genetic ancestry.

332
00:16:25,300 --> 00:16:27,470
Think about if anybody's tried 23 in me,

333
00:16:27,470 --> 00:16:31,120
it's similar to the kind of
results you get from that.

334
00:16:31,120 --> 00:16:34,173
And what we could see is
here, are non Africans,

335
00:16:35,010 --> 00:16:37,030
people who self-identify as European

336
00:16:37,030 --> 00:16:39,003
or middle Eastern are shown in blue.

337
00:16:40,620 --> 00:16:44,280
People from India shown here
in the sort of fissure color,

338
00:16:44,280 --> 00:16:47,790
and then we have people
from Pakistan, central Asia,

339
00:16:47,790 --> 00:16:51,340
East Asia, Oceania, and the Americas.

340
00:16:51,340 --> 00:16:52,510
When we look at Africa,

341
00:16:52,510 --> 00:16:56,470
you see a lot of colors relative
to the rest of the world.

342
00:16:56,470 --> 00:16:57,440
And what that's telling you

343
00:16:57,440 --> 00:16:59,500
is there's a lot of genetic diversity

344
00:16:59,500 --> 00:17:02,390
and a lot of what we call
population substructure

345
00:17:02,390 --> 00:17:06,740
or genetic differentiation
amongst the populations in Africa

346
00:17:06,740 --> 00:17:08,440
compared to the rest of the world.

347
00:17:09,550 --> 00:17:12,510
Now, if we repeat this
analysis just for Africans

348
00:17:12,510 --> 00:17:16,840
and we pool individuals by
geographic region or language,

349
00:17:16,840 --> 00:17:19,770
again, you could see all
the diversity in Africa

350
00:17:19,770 --> 00:17:23,850
and that populations in Western
and Central and Eastern,

351
00:17:23,850 --> 00:17:25,740
Southern and Northern Africa,

352
00:17:25,740 --> 00:17:27,800
are very divergent from each other,

353
00:17:27,800 --> 00:17:29,490
and that even within a region,

354
00:17:29,490 --> 00:17:33,470
there's so much diversity,
so much admixture.

355
00:17:33,470 --> 00:17:35,310
You can also look at these faces

356
00:17:35,310 --> 00:17:39,090
and see that there's a lot
of phenotypic variation.

357
00:17:39,090 --> 00:17:41,130
And we think that this is reflecting

358
00:17:41,130 --> 00:17:43,740
the demographic history of Africans,

359
00:17:43,740 --> 00:17:47,073
and also their adaptation
to different environments.

360
00:17:49,820 --> 00:17:52,860
So, in terms of the
African American ancestry,

361
00:17:52,860 --> 00:17:53,920
not surprisingly,

362
00:17:53,920 --> 00:17:57,621
we see predominantly
West African ancestry.

363
00:17:57,621 --> 00:18:02,621
We see some European ancestry
on average, it's around 20%.

364
00:18:02,790 --> 00:18:05,655
We see small amounts of
ancestry from other groups

365
00:18:05,655 --> 00:18:08,550
and a little bit from this
Nilo-Saharan ancestry,

366
00:18:08,550 --> 00:18:10,070
which I think is actually representing

367
00:18:10,070 --> 00:18:11,600
a group called the Fulani,

368
00:18:11,600 --> 00:18:14,430
who are very common throughout

369
00:18:14,430 --> 00:18:17,083
sort of the Sahel region in Africa.

370
00:18:18,430 --> 00:18:20,290
Again, this is simply reflecting

371
00:18:20,290 --> 00:18:22,750
the history of the slave trade.

372
00:18:22,750 --> 00:18:24,700
Now, one of the biggest sources

373
00:18:24,700 --> 00:18:26,930
of the slave trade in fact was Angola.

374
00:18:26,930 --> 00:18:29,670
And I'll just point out that
we currently know very little

375
00:18:29,670 --> 00:18:31,050
about variation in that region.

376
00:18:31,050 --> 00:18:32,500
It's been a place that,

377
00:18:32,500 --> 00:18:35,310
it's been dangerous, frankly,
to do research there.

378
00:18:35,310 --> 00:18:38,300
So there's a lot we still
have to learn about,

379
00:18:38,300 --> 00:18:41,250
in terms of both African history,

380
00:18:41,250 --> 00:18:44,423
and African American ancestry.

381
00:18:46,410 --> 00:18:48,100
So I now wanna tell you about a study

382
00:18:48,100 --> 00:18:50,800
that was done by a former
post-doc Shauhua Fan,

383
00:18:50,800 --> 00:18:51,930
in collaboration with

384
00:18:51,930 --> 00:18:55,370
the Simons Genome Diversity
Project and David Reich,

385
00:18:55,370 --> 00:19:00,370
and we sequenced 94 individuals
from 44 African populations

386
00:19:01,090 --> 00:19:02,560
at the whole genome levels.

387
00:19:02,560 --> 00:19:05,070
So sequencing the entire genome,

388
00:19:05,070 --> 00:19:08,270
and from that, we could construct again,

389
00:19:08,270 --> 00:19:09,780
one of these phylogenetic trees

390
00:19:09,780 --> 00:19:11,960
and it's consistent with
what I showed you before,

391
00:19:11,960 --> 00:19:15,640
the San have the most ancestral lineages,

392
00:19:15,640 --> 00:19:16,730
followed by the Pygmies,

393
00:19:16,730 --> 00:19:19,950
and then we see clustering
mainly by geographic region,

394
00:19:19,950 --> 00:19:21,483
non-African would be up here.

395
00:19:23,150 --> 00:19:26,610
We can also use computational
approaches to make inferences

396
00:19:26,610 --> 00:19:29,520
about changes in the
effect of population size,

397
00:19:29,520 --> 00:19:31,400
going backwards in time.

398
00:19:31,400 --> 00:19:32,370
And what's interesting.

399
00:19:32,370 --> 00:19:34,990
You can just ignore all the
stuff that's really recent

400
00:19:34,990 --> 00:19:38,060
because this method doesn't
do very well for that.

401
00:19:38,060 --> 00:19:40,120
But we could see that we
start seeing divergence

402
00:19:40,120 --> 00:19:42,190
around 250,000 years ago,

403
00:19:42,190 --> 00:19:45,400
close to the fossil evidence

404
00:19:45,400 --> 00:19:46,910
for the origin of modern humans.

405
00:19:46,910 --> 00:19:48,300
And that in the past,

406
00:19:48,300 --> 00:19:50,040
the San in the Pygmies have maintained

407
00:19:50,040 --> 00:19:52,540
the largest effect of population size,

408
00:19:52,540 --> 00:19:56,003
even though today, they have
a relative small census size.

409
00:19:57,220 --> 00:19:58,410
We can also estimate

410
00:19:58,410 --> 00:20:00,600
when populations diverge from each other.

411
00:20:00,600 --> 00:20:03,510
And we see that the earliest
divergence is the San

412
00:20:03,510 --> 00:20:04,620
from all other groups,

413
00:20:04,620 --> 00:20:08,120
could have been as early
as 160,000 years ago.

414
00:20:08,120 --> 00:20:09,840
We see divergence between the San

415
00:20:09,840 --> 00:20:11,683
and some of these other
hunter-gatherer groups

416
00:20:11,683 --> 00:20:14,230
like the Hadza, the
Sandawe and the Pygmies,

417
00:20:14,230 --> 00:20:17,570
around 70 to 85,000 years ago.

418
00:20:17,570 --> 00:20:20,760
We see divergence between
Niger-Kordofanian speakers,

419
00:20:20,760 --> 00:20:23,960
Afro-Asiatic, Nilo-Saharan speakers,

420
00:20:23,960 --> 00:20:28,960
around 16,000 to roughly 35,000 years ago.

421
00:20:29,680 --> 00:20:32,030
But even amongst some
of these San populations

422
00:20:32,030 --> 00:20:34,580
that speak slightly divergent languages,

423
00:20:34,580 --> 00:20:36,700
they are estimated to have diverged

424
00:20:36,700 --> 00:20:38,790
as much as 30,000 years ago.

425
00:20:38,790 --> 00:20:40,710
So again, it's consistent
with this history

426
00:20:40,710 --> 00:20:43,870
of sub-structure populations,
having been separated

427
00:20:43,870 --> 00:20:45,600
for a long period of time,

428
00:20:45,600 --> 00:20:50,600
but with a lot of migration
and admixture between groups.

429
00:20:52,170 --> 00:20:53,390
I wanna tell you just a little bit

430
00:20:53,390 --> 00:20:55,160
about phenotypic variation

431
00:20:55,160 --> 00:20:58,040
for some of the traits that we've studied.

432
00:20:58,040 --> 00:21:01,370
And this is work that was
done by Matt Hansen in my lab.

433
00:21:01,370 --> 00:21:05,240
And we are looking at groups
that have very different diets.

434
00:21:05,240 --> 00:21:07,360
So hunter-gatherers have
a diet that's heavy,

435
00:21:07,360 --> 00:21:12,170
in nuts and tubers and fruits
to a lesser extent, meat,

436
00:21:12,170 --> 00:21:14,730
pastoralists have a diet
heavy in milk and blood,

437
00:21:14,730 --> 00:21:16,730
and to a lesser extent meat

438
00:21:16,730 --> 00:21:18,330
and the agriculturalists have a diet

439
00:21:18,330 --> 00:21:22,733
heavy in grain and maize,
fruits, and tubers.

440
00:21:23,690 --> 00:21:26,860
So here we have,

441
00:21:26,860 --> 00:21:28,190
these are different populations

442
00:21:28,190 --> 00:21:31,570
we're actually just looking
at the data from women here,

443
00:21:31,570 --> 00:21:33,360
and they've been color coded according

444
00:21:33,360 --> 00:21:35,780
to the subsistence pattern.

445
00:21:35,780 --> 00:21:37,467
And we could see if we look
at height, for example,

446
00:21:37,467 --> 00:21:41,160
at the very, the short end

447
00:21:41,160 --> 00:21:43,870
are these so-called pygmy populations,

448
00:21:43,870 --> 00:21:47,047
and at the very high end are
some of the pastoralist groups

449
00:21:47,047 --> 00:21:50,380
and some of the food
producing agriculturalists.

450
00:21:50,380 --> 00:21:51,920
At the low end for BMI,

451
00:21:51,920 --> 00:21:53,900
are some of the East African pastoralists.

452
00:21:53,900 --> 00:21:57,210
They tend to have very
tall and thin body types.

453
00:21:57,210 --> 00:21:59,773
And at the high end are some
of the agriculturalists.

454
00:22:02,362 --> 00:22:03,940
If we look at waist circumference,

455
00:22:03,940 --> 00:22:05,180
we could see at the very low end

456
00:22:05,180 --> 00:22:07,560
are actually the San populations

457
00:22:07,560 --> 00:22:12,400
and at the higher end are some
of the food producing groups,

458
00:22:12,400 --> 00:22:16,450
and percent body fat, we see a
similar trend at the high end

459
00:22:16,450 --> 00:22:19,070
or the food producing
groups in the Herero.

460
00:22:19,070 --> 00:22:21,490
So this is a Herero woman.

461
00:22:21,490 --> 00:22:24,270
This ethnic group lives
in Southern Africa.

462
00:22:24,270 --> 00:22:26,050
They speak a language related

463
00:22:26,050 --> 00:22:28,540
to these neighboring
agriculturalist groups, and in fact,

464
00:22:28,540 --> 00:22:31,560
they came from that Bantu
migration that I mentioned.

465
00:22:31,560 --> 00:22:34,000
But they consider it culturally desirable

466
00:22:34,000 --> 00:22:35,630
for women to be very heavy.

467
00:22:35,630 --> 00:22:37,730
They see it as a sign of fecundity.

468
00:22:37,730 --> 00:22:40,963
So I wasn't an entirely
surprised to see this result.

469
00:22:42,270 --> 00:22:44,380
So now for the last part of my talk,

470
00:22:44,380 --> 00:22:46,790
I wanna talk about our studies

471
00:22:46,790 --> 00:22:48,897
of the genetics of adaptation.

472
00:22:48,897 --> 00:22:52,957
And I love this quote from
Charles Darwin's classic book,

473
00:22:52,957 --> 00:22:54,960
"On the Origin of Species",

474
00:22:54,960 --> 00:22:58,817
and he says, "This preservation
of favorable variations

475
00:22:58,817 --> 00:23:01,547
"and the rejection of
injurious variations,

476
00:23:01,547 --> 00:23:03,587
"I call natural selection."

477
00:23:05,850 --> 00:23:06,890
So what I'm showing here

478
00:23:06,890 --> 00:23:11,630
are different examples of
how people have adapted

479
00:23:11,630 --> 00:23:13,010
to different environments

480
00:23:13,010 --> 00:23:17,200
and different diets across the globe.

481
00:23:17,200 --> 00:23:20,810
You might have heard for
example, of Sickle Cell Disease,

482
00:23:20,810 --> 00:23:22,690
being common in people from Africa,

483
00:23:22,690 --> 00:23:24,070
where there's a lot of malaria

484
00:23:24,070 --> 00:23:27,410
because people who have one
copy of a sickle mutation,

485
00:23:27,410 --> 00:23:29,180
and one of the normal copies,

486
00:23:29,180 --> 00:23:31,720
are actually protected from malaria.

487
00:23:31,720 --> 00:23:34,880
So this variant remains in the population.

488
00:23:34,880 --> 00:23:36,900
But the example I'm gonna
tell you about right now

489
00:23:36,900 --> 00:23:39,220
is the genetics of lactose tolerance.

490
00:23:39,220 --> 00:23:41,523
The ability to drink milk as adults.

491
00:23:42,682 --> 00:23:47,682
So the ability to drink milk
is due to the expression

492
00:23:48,620 --> 00:23:50,650
of a gene that codes for lactase,

493
00:23:50,650 --> 00:23:52,863
actually Lactase-phlorizin hydrolase,

494
00:23:53,713 --> 00:23:57,230
and this enzyme breaks
down the sugar lactose

495
00:23:57,230 --> 00:23:59,010
into glucose and galactose,

496
00:23:59,010 --> 00:24:01,670
and those are rapidly taken
up into the bloodstream

497
00:24:01,670 --> 00:24:03,530
in people who can digest milk.

498
00:24:05,305 --> 00:24:09,353
But in people who can't digest
milk shortly after weaning,

499
00:24:11,070 --> 00:24:14,170
we can see that they are not able,

500
00:24:14,170 --> 00:24:16,060
they're not expressing this,

501
00:24:16,060 --> 00:24:18,550
they're no longer expressing
lactase after weaning,

502
00:24:18,550 --> 00:24:20,860
and so they're not able
to break down that sugar.

503
00:24:20,860 --> 00:24:22,210
It's gonna go to the lower guts,

504
00:24:22,210 --> 00:24:25,750
it's gonna cause severe
intestinal distress.

505
00:24:25,750 --> 00:24:27,840
And interestingly, I mean,

506
00:24:27,840 --> 00:24:30,810
anthropologists have
noted for quite some time,

507
00:24:30,810 --> 00:24:35,810
that the ancestral state is
actually lactoses intolerance

508
00:24:36,430 --> 00:24:39,080
and lactoses tolerance,
the ability to drink milk

509
00:24:39,080 --> 00:24:41,360
as an adult is a drive state.

510
00:24:41,360 --> 00:24:43,210
It's an adaptation.

511
00:24:43,210 --> 00:24:44,670
And we see that it's most common

512
00:24:44,670 --> 00:24:47,250
in Northern European population.

513
00:24:47,250 --> 00:24:51,160
It's a little bit less
common in Southern Europeans.

514
00:24:51,160 --> 00:24:54,060
It's very uncommon in
East Asian populations.

515
00:24:54,060 --> 00:24:58,130
And it's uncommon in
most African populations,

516
00:24:58,130 --> 00:25:02,033
except for those who practice
during who are pastoralists.

517
00:25:03,500 --> 00:25:07,130
So we wanted to understand the
genetic basis of this trait.

518
00:25:07,130 --> 00:25:10,010
And so starting in, I think it was 2002,

519
00:25:12,370 --> 00:25:14,890
we gave what is called a
lactose tolerance test.

520
00:25:14,890 --> 00:25:17,960
So giving the sugar lactose,

521
00:25:17,960 --> 00:25:21,840
and then every 20 minutes
measuring the blood glucose

522
00:25:21,840 --> 00:25:24,543
using like a diabetes monitoring kit.

523
00:25:26,520 --> 00:25:27,470
And as I said,

524
00:25:27,470 --> 00:25:29,580
this is every 20 minutes over a period

525
00:25:29,580 --> 00:25:30,893
of a little over an hour,

526
00:25:32,520 --> 00:25:35,930
and then we look at the maximum
rise in the blood sugar.

527
00:25:35,930 --> 00:25:38,290
And according to the medical literature,

528
00:25:38,290 --> 00:25:41,970
if you have a rise that's
greater than 1.7 millimolar,

529
00:25:41,970 --> 00:25:43,530
then you're lactose tolerant,

530
00:25:43,530 --> 00:25:46,380
if it's less than 1.1,
you're lactose intolerant,

531
00:25:46,380 --> 00:25:48,430
and some people are something in-between.

532
00:25:51,250 --> 00:25:53,850
Now around that time,

533
00:25:53,850 --> 00:25:58,460
there was a beautiful study
done by a group in Finland

534
00:25:58,460 --> 00:26:00,220
that had identified a mutation

535
00:26:00,220 --> 00:26:05,060
associated with lactose tolerance
in European populations.

536
00:26:05,060 --> 00:26:08,870
And it was in an intron
and non-coding region

537
00:26:08,870 --> 00:26:13,750
of this neighboring gene
and a regulatory region.

538
00:26:13,750 --> 00:26:16,350
So we sequenced that
region in the Africans

539
00:26:16,350 --> 00:26:17,260
and we didn't see it.

540
00:26:17,260 --> 00:26:20,730
We didn't see the same mutation
that's found in Europe,

541
00:26:20,730 --> 00:26:24,100
but we saw three other
variants that were associated

542
00:26:24,100 --> 00:26:26,790
with lactose tolerance, and as I said,

543
00:26:26,790 --> 00:26:30,270
they arose independently
from the European mutation

544
00:26:30,270 --> 00:26:33,120
and they have an interesting
geographic distribution.

545
00:26:33,120 --> 00:26:35,160
The common variant in East Africa

546
00:26:35,160 --> 00:26:37,553
is at position minus 14010.

547
00:26:39,020 --> 00:26:41,120
This variant in the Middle East

548
00:26:41,120 --> 00:26:43,670
is at position 13915,

549
00:26:43,670 --> 00:26:46,800
and we think it was introduced
into Africa by migration.

550
00:26:46,800 --> 00:26:50,110
And then this variant at position 13907,

551
00:26:50,110 --> 00:26:51,963
is common in the Horn of Africa.

552
00:26:54,500 --> 00:26:57,220
We also wanted to see if
there was a genomic signature

553
00:26:57,220 --> 00:26:59,000
of natural selection.

554
00:26:59,000 --> 00:27:00,400
So we use something called

555
00:27:00,400 --> 00:27:02,900
an extended haplotype homozygosity test.

556
00:27:02,900 --> 00:27:03,733
What is that?

557
00:27:03,733 --> 00:27:07,800
Well, basically, this shows
you what the process looks like

558
00:27:07,800 --> 00:27:08,910
or what you might expect

559
00:27:08,910 --> 00:27:11,490
under a very strong recent selection.

560
00:27:11,490 --> 00:27:13,330
So here are some, this is,

561
00:27:13,330 --> 00:27:15,330
let's say it's around the lactase gene

562
00:27:15,330 --> 00:27:16,790
or the regulatory region,

563
00:27:16,790 --> 00:27:19,700
and there's some variation in that region,

564
00:27:19,700 --> 00:27:22,440
and a mutation arises, and in this case,

565
00:27:22,440 --> 00:27:25,410
it's the variant that's
associated with lactose tolerance.

566
00:27:25,410 --> 00:27:29,560
And if it increases the
fitness of the individuals,

567
00:27:29,560 --> 00:27:30,780
so they have more children,

568
00:27:30,780 --> 00:27:32,660
and their children have more children,

569
00:27:32,660 --> 00:27:34,900
it's going to rapidly
rise to high frequency

570
00:27:34,900 --> 00:27:37,610
in the population, as shown here.

571
00:27:37,610 --> 00:27:39,970
And when that happens,
it's dragging with it,

572
00:27:39,970 --> 00:27:41,650
the neighboring variation,

573
00:27:41,650 --> 00:27:45,363
and that actually can go on
for millions of nucleotides.

574
00:27:46,590 --> 00:27:48,540
And so this is what we see,

575
00:27:48,540 --> 00:27:52,250
is that in red I'm showing
individuals who are homozygous.

576
00:27:52,250 --> 00:27:54,960
They have two copies of the C variant

577
00:27:54,960 --> 00:27:58,070
that is associated with lactose tolerance.

578
00:27:58,070 --> 00:28:00,580
And we could see that there
are homozygous going out,

579
00:28:00,580 --> 00:28:03,210
almost 3,000,000 nucleotides.

580
00:28:03,210 --> 00:28:04,800
If we compare that to people

581
00:28:04,800 --> 00:28:07,430
who have the ancestral gene variant,

582
00:28:07,430 --> 00:28:09,890
we could see that they
don't have this kind of

583
00:28:09,890 --> 00:28:13,520
extended homozygosity that
it only extends on average,

584
00:28:13,520 --> 00:28:15,780
about 1800 base pairs.

585
00:28:15,780 --> 00:28:17,900
Now, one of the ways you
can estimate the age of this

586
00:28:17,900 --> 00:28:21,230
is to look at the breakdown of this,

587
00:28:21,230 --> 00:28:24,310
basically how these red
lines are breaking down

588
00:28:24,310 --> 00:28:26,990
due to a combination, every generation.

589
00:28:26,990 --> 00:28:28,920
And using computational approaches,

590
00:28:28,920 --> 00:28:32,300
we can estimate the age of
this East African variant

591
00:28:32,300 --> 00:28:35,620
to be around 3,000 to 7,000 years old.

592
00:28:35,620 --> 00:28:37,760
And that's interesting
because it corresponds

593
00:28:37,760 --> 00:28:40,000
with the origin of cattle domestication

594
00:28:40,000 --> 00:28:41,560
based on the archeological record,

595
00:28:41,560 --> 00:28:44,530
it's about to have
originated in either Sudan,

596
00:28:44,530 --> 00:28:47,890
or the Middle East at around
nine to 10,000 years ago,

597
00:28:47,890 --> 00:28:49,310
which by the way we estimated

598
00:28:49,310 --> 00:28:52,230
was the age of the European mutation,

599
00:28:52,230 --> 00:28:57,230
but cattle domestication
was not introduced

600
00:28:57,610 --> 00:29:01,410
south of the Saharan Desert
till roughly 5,000 years ago.

601
00:29:01,410 --> 00:29:03,730
And so it corresponds really well

602
00:29:03,730 --> 00:29:06,900
with the age estimates that we obtained.

603
00:29:06,900 --> 00:29:10,563
And it's a great example of
gene culture co-evolution.

604
00:29:12,140 --> 00:29:14,450
So lastly, I wanna tell
you about our study

605
00:29:14,450 --> 00:29:18,313
of the genetics of skin pigmentation.

606
00:29:20,200 --> 00:29:24,070
So it's thought that skin
color is an adaptive trait,

607
00:29:24,070 --> 00:29:26,390
in the upper left, these blue dots

608
00:29:26,390 --> 00:29:28,690
are representing the levels of melanin,

609
00:29:28,690 --> 00:29:31,500
that's the dark pigment in skin.

610
00:29:31,500 --> 00:29:35,090
And you could see that it is
correlated with UV levels.

611
00:29:35,090 --> 00:29:35,940
And it's thought that

612
00:29:35,940 --> 00:29:39,230
when our ancestors migrated out of Africa,

613
00:29:39,230 --> 00:29:42,260
that there would've been
selection for lighter skin,

614
00:29:42,260 --> 00:29:45,130
possibly to promote
synthesis of vitamin D,

615
00:29:45,130 --> 00:29:49,950
it's synthesized in the
skin in response to UV,

616
00:29:49,950 --> 00:29:52,880
and in places where people
are close to the equator

617
00:29:52,880 --> 00:29:54,620
and where there's a lot of UV,

618
00:29:54,620 --> 00:29:58,080
there'd be selection for darker
skin to prevent skin cancer

619
00:29:58,080 --> 00:30:01,990
and possibly break down a folic acid,

620
00:30:01,990 --> 00:30:05,410
which is very important for
development of the fetus.

621
00:30:05,410 --> 00:30:08,030
So most studies have
been done in Europeans,

622
00:30:08,030 --> 00:30:10,670
the study of this trait in Africans,

623
00:30:10,670 --> 00:30:12,670
we use a spectrophotometer,

624
00:30:12,670 --> 00:30:14,820
we shine the light under the arm,

625
00:30:14,820 --> 00:30:18,060
that's an area that's not
getting a lot of sunlight,

626
00:30:18,060 --> 00:30:22,160
and we measure the
wavelength of the reflection.

627
00:30:22,160 --> 00:30:25,710
And from that we can
infer the melanin levels.

628
00:30:25,710 --> 00:30:29,840
And what we see is a big
range of variation in Africa.

629
00:30:29,840 --> 00:30:32,810
The most lightly pigmented
population are the San.

630
00:30:32,810 --> 00:30:33,870
And remember I told you

631
00:30:33,870 --> 00:30:36,763
that they have the
oldest genetic lineages,

632
00:30:38,607 --> 00:30:40,710
and then we see a lot of variation

633
00:30:40,710 --> 00:30:43,350
with the most darkly pigmented people

634
00:30:43,350 --> 00:30:46,110
being the Nilo-Saharan
speaking pastoralists

635
00:30:46,110 --> 00:30:48,283
who originated in South Sudan.

636
00:30:50,810 --> 00:30:52,910
Then we did something that is called

637
00:30:52,910 --> 00:30:54,810
a Genome-Wide Association Study.

638
00:30:54,810 --> 00:30:59,080
So what we do is we look at
single nucleotide variants,

639
00:30:59,080 --> 00:31:01,420
variable parts of the genome.

640
00:31:01,420 --> 00:31:04,220
We look at millions of them,

641
00:31:04,220 --> 00:31:07,310
in this case, 4,000,000
sites of the genome,

642
00:31:07,310 --> 00:31:10,390
and we look for an
association with skin color.

643
00:31:10,390 --> 00:31:12,680
And if there's an association,

644
00:31:12,680 --> 00:31:15,870
then it means that that variable
site that you're looking at

645
00:31:15,870 --> 00:31:20,593
is near a causal mutation that
is impacting pigmentation.

646
00:31:21,740 --> 00:31:25,500
So we found eight new sites
at four regions of the genome

647
00:31:25,500 --> 00:31:27,780
shown here, and I'm gonna
just quickly step through.

648
00:31:27,780 --> 00:31:30,720
And this is work that
was done by Nick Crawford

649
00:31:30,720 --> 00:31:32,023
and Matt Hansen in my lab.

650
00:31:33,710 --> 00:31:38,063
So the strongest association
was at a gene called SLC24A5,

651
00:31:38,907 --> 00:31:40,020
(coughs) excuse me,

652
00:31:40,020 --> 00:31:43,520
there's an amino acid
substitution in this gene

653
00:31:43,520 --> 00:31:46,500
that actually, this was
already known to play a role

654
00:31:46,500 --> 00:31:48,260
in light-skin color in Europeans.

655
00:31:48,260 --> 00:31:51,540
It was actually the very
first gene identified

656
00:31:51,540 --> 00:31:53,803
to play a role in skin color in humans.

657
00:31:54,760 --> 00:31:57,415
It was first identified, I
believe, in a zebra fish.

658
00:31:57,415 --> 00:31:59,209
And then they looked at the distribution

659
00:31:59,209 --> 00:32:00,482
in human populations,

660
00:32:00,482 --> 00:32:01,940
and we could see that the variant

661
00:32:01,940 --> 00:32:04,900
associated with light skin, shown in blue,

662
00:32:04,900 --> 00:32:08,320
is at almost 100% frequency in Europe.

663
00:32:08,320 --> 00:32:12,000
It's also really common in
Pakistan, in some parts of India.

664
00:32:12,000 --> 00:32:14,970
And you could see it's
really common in East Africa.

665
00:32:14,970 --> 00:32:16,370
So we wanted to know,

666
00:32:16,370 --> 00:32:20,220
did that mutation arise
independently in Africa,

667
00:32:20,220 --> 00:32:22,193
or was it introduced by migration?

668
00:32:23,070 --> 00:32:24,760
So one of the ways we can do this

669
00:32:24,760 --> 00:32:27,180
is to construct something
called a haplotype network.

670
00:32:27,180 --> 00:32:28,938
So what's a haplotype?

671
00:32:28,938 --> 00:32:32,600
A haplotype is just how the
different variants are arranged

672
00:32:32,600 --> 00:32:35,350
along some stretch of the chromosome.

673
00:32:35,350 --> 00:32:39,160
In this case, we're looking
over 70,000 nucleotides.

674
00:32:39,160 --> 00:32:41,460
Each circle represents a haplotype.

675
00:32:41,460 --> 00:32:44,060
The size of the circle
indicates the number of people

676
00:32:44,060 --> 00:32:45,830
who have it, and the colors represent

677
00:32:45,830 --> 00:32:49,270
the relative proportion
in different populations.

678
00:32:49,270 --> 00:32:51,880
This is also kinda like a
phylogenetic tree in that

679
00:32:51,880 --> 00:32:55,060
these haplotypes are very
different genetically

680
00:32:55,060 --> 00:32:56,940
from these haplotypes.

681
00:32:56,940 --> 00:33:00,020
And what we see is that this variant

682
00:33:00,020 --> 00:33:01,390
associated with light skin color

683
00:33:01,390 --> 00:33:03,610
is really common in Europeans,

684
00:33:03,610 --> 00:33:05,830
and it's on this long
haplotype background,

685
00:33:05,830 --> 00:33:08,763
which is consistent with
recent positive selection.

686
00:33:10,540 --> 00:33:12,390
The East Africans have that variant

687
00:33:12,390 --> 00:33:15,970
on the identical chromosomal background.

688
00:33:15,970 --> 00:33:18,870
So what that tells us is that
this was likely introduced

689
00:33:18,870 --> 00:33:21,420
by back migration into Africa.

690
00:33:21,420 --> 00:33:24,793
We think within the
past 5,000 or so years.

691
00:33:26,940 --> 00:33:29,470
Now, the second strongest
association was at a gene

692
00:33:29,470 --> 00:33:32,690
that had no name at the
time that we studied this.

693
00:33:32,690 --> 00:33:34,773
We're the first people to
characterize this gene.

694
00:33:34,773 --> 00:33:37,540
It's called MFSD12.

695
00:33:37,540 --> 00:33:41,520
And we found two different associations,

696
00:33:41,520 --> 00:33:43,380
one in a regulatory region,

697
00:33:43,380 --> 00:33:46,450
it's regulating the
expression of this gene.

698
00:33:46,450 --> 00:33:48,340
And the other is in
the coding region here.

699
00:33:48,340 --> 00:33:50,420
We don't know exactly what it's doing.

700
00:33:50,420 --> 00:33:53,120
It could be altering splicing of this gene

701
00:33:53,120 --> 00:33:55,720
or something else that
we still don't know.

702
00:33:55,720 --> 00:33:57,300
So if this variant upstream,

703
00:33:57,300 --> 00:34:00,250
we could see that the variant associated

704
00:34:00,250 --> 00:34:04,780
with light skin color is
common in Europe, in East Asia

705
00:34:04,780 --> 00:34:09,610
and in India, in East
Africa, and look at the San,

706
00:34:09,610 --> 00:34:12,333
how common it is in the
lightly pigmented San.

707
00:34:13,369 --> 00:34:15,410
So we did a number of functional studies

708
00:34:15,410 --> 00:34:17,540
to characterize this and we
worked with our collaborator,

709
00:34:17,540 --> 00:34:19,370
William Pavan at NIH,

710
00:34:19,370 --> 00:34:20,677
and he knocked this gene out using

711
00:34:20,677 --> 00:34:24,460
CRISPR-Cas9 gene editing technology,

712
00:34:24,460 --> 00:34:26,780
and it had a pretty dramatic impact

713
00:34:26,780 --> 00:34:28,100
on the color of this mouse.

714
00:34:28,100 --> 00:34:29,370
This isn't a goodie mouse.

715
00:34:29,370 --> 00:34:33,330
It has a sort of
brownish-yellowish sort of color

716
00:34:33,330 --> 00:34:35,430
that's actually produced by pheomelanin.

717
00:34:35,430 --> 00:34:38,920
That's the pigment that gives
sort of a yellow, brown color.

718
00:34:38,920 --> 00:34:40,280
And when it's knocked out,

719
00:34:40,280 --> 00:34:42,993
you see that this underlying gray color.

720
00:34:44,540 --> 00:34:47,320
So this gene has now been
characterized by others.

721
00:34:47,320 --> 00:34:50,600
It's been shown to play a role
in the color of, I believe,

722
00:34:50,600 --> 00:34:54,350
ponies and dogs and
different human populations,

723
00:34:54,350 --> 00:34:56,230
and has been shown to play a role

724
00:34:56,230 --> 00:34:58,683
in risk for melanoma skin cancer.

725
00:35:00,710 --> 00:35:03,870
Then the third association
where two different sites near

726
00:35:03,870 --> 00:35:07,790
a gene called DDB1, and
what does that gene do?

727
00:35:07,790 --> 00:35:10,620
Well, that gene plays
a really important role

728
00:35:10,620 --> 00:35:15,620
in repairing DNA when it's
damaged by UV exposure.

729
00:35:15,970 --> 00:35:17,320
So that's kind of interesting,

730
00:35:17,320 --> 00:35:19,850
'cause there's like this UV connection,

731
00:35:19,850 --> 00:35:21,090
but what does that have to do,

732
00:35:21,090 --> 00:35:22,950
and I was gonna say that,

733
00:35:22,950 --> 00:35:25,670
people who have a mutation in that gene

734
00:35:25,670 --> 00:35:28,220
or in genes that code for proteins

735
00:35:28,220 --> 00:35:30,640
that interact with that gene,

736
00:35:30,640 --> 00:35:34,300
have a disease called
Xeroderma Pigmentosum.

737
00:35:34,300 --> 00:35:36,850
They're not able to be exposed to sunlight

738
00:35:36,850 --> 00:35:41,150
because they can't repair
the damage to their DNA.

739
00:35:41,150 --> 00:35:42,370
But again, what does this

740
00:35:42,370 --> 00:35:44,410
have to do with skin color in humans?

741
00:35:44,410 --> 00:35:45,250
Well, we don't know,

742
00:35:45,250 --> 00:35:46,270
but the interesting thing is,

743
00:35:46,270 --> 00:35:48,410
when I was looking this gene up,

744
00:35:48,410 --> 00:35:50,400
it turns out that it is the gene

745
00:35:50,400 --> 00:35:53,150
that causes the color of tomatoes.

746
00:35:53,150 --> 00:35:55,620
So it is playing a role in pigmentation.

747
00:35:55,620 --> 00:35:58,220
We just don't know exactly
what it is yet in humans.

748
00:35:59,220 --> 00:36:03,340
It also showed a very strong
signature of a selective sweep.

749
00:36:03,340 --> 00:36:05,930
So when this variant, the light variant,

750
00:36:05,930 --> 00:36:07,950
was introduced out of Africa,

751
00:36:07,950 --> 00:36:12,210
it rapidly rose to high
frequency and those haplotypes,

752
00:36:12,210 --> 00:36:15,480
it drags with it all the
neighboring variation.

753
00:36:15,480 --> 00:36:18,413
And so what you see is actually
a decrease in variation.

754
00:36:18,413 --> 00:36:22,150
This is an indication of
how variable a site is.

755
00:36:22,150 --> 00:36:25,420
And you see a decrease of
variation in Asians and Europeans

756
00:36:25,420 --> 00:36:30,270
across a very large region,
it's about 500,000 nucleotides.

757
00:36:30,270 --> 00:36:33,083
We don't really see these
in Africans so much.

758
00:36:34,480 --> 00:36:36,390
And when you look at
the haplotype network,

759
00:36:36,390 --> 00:36:39,050
we see this long, common haplotype.

760
00:36:39,050 --> 00:36:42,540
This is about 200,000 nucleotides,

761
00:36:42,540 --> 00:36:46,565
and very common only in
non-African populations.

762
00:36:46,565 --> 00:36:48,773
So here would be the African populations.

763
00:36:50,520 --> 00:36:54,130
We can then construct what's
called a gene geneology.

764
00:36:54,130 --> 00:36:56,930
So think of this, if we unwounded,

765
00:36:56,930 --> 00:36:59,750
it's like that phylogeny
that I showed you,

766
00:36:59,750 --> 00:37:01,880
and at the end of each of these branches

767
00:37:01,880 --> 00:37:05,200
is actually an individual
sequence at that region.

768
00:37:05,200 --> 00:37:07,693
And if the individual, if at that region,

769
00:37:07,693 --> 00:37:10,010
they have the variant
associated with light skin,

770
00:37:10,010 --> 00:37:11,700
they have this open circle,

771
00:37:11,700 --> 00:37:13,510
and they have the variant dark skin,

772
00:37:13,510 --> 00:37:15,570
they're shown over here,

773
00:37:15,570 --> 00:37:19,750
and in this particular case,

774
00:37:19,750 --> 00:37:24,750
we could see evidence that
there was a rapid coalescence

775
00:37:24,800 --> 00:37:26,750
around 60,000 years ago,

776
00:37:26,750 --> 00:37:30,910
which is around the time
of migration out of Africa.

777
00:37:30,910 --> 00:37:34,900
So this is very consistent
with the rapid selective sweep

778
00:37:34,900 --> 00:37:37,500
occurring when modern humans left Africa,

779
00:37:37,500 --> 00:37:39,500
we still don't know the reason for that.

780
00:37:41,230 --> 00:37:42,560
And then the last region

781
00:37:42,560 --> 00:37:46,333
was near two genes called OCA2 and HERC2.

782
00:37:47,920 --> 00:37:50,090
Now these were already
known to play a role

783
00:37:50,090 --> 00:37:51,560
in skin pigmentation.

784
00:37:51,560 --> 00:37:53,950
And actually there had been
mutations found in this,

785
00:37:53,950 --> 00:37:58,500
HERC2 gene that are enhancers
or regulatory regions

786
00:37:58,500 --> 00:38:01,930
that influence the expression of OCA2.

787
00:38:01,930 --> 00:38:04,210
And in non-Africans,

788
00:38:04,210 --> 00:38:07,513
they play a role in light
skin color and eye color.

789
00:38:08,350 --> 00:38:10,730
But the mutations that we found

790
00:38:10,730 --> 00:38:13,470
arose independently from those,

791
00:38:13,470 --> 00:38:16,230
but we did show that they are influencing

792
00:38:16,230 --> 00:38:18,003
the expression of OCA2.

793
00:38:19,260 --> 00:38:21,520
Now, then we found this other variant

794
00:38:21,520 --> 00:38:23,910
that is what's called
a synonymous variant.

795
00:38:23,910 --> 00:38:27,770
It's not causing an amino
acid change in Exon 10

796
00:38:27,770 --> 00:38:30,803
in the coding region of this OCA2 gene.

797
00:38:32,120 --> 00:38:34,510
Now it happened to be in a
region that had previously

798
00:38:34,510 --> 00:38:39,040
been shown to cause certain mutations

799
00:38:39,040 --> 00:38:41,033
in that region, cause albinism,

800
00:38:42,430 --> 00:38:44,910
but it wasn't associated
with gene expression.

801
00:38:44,910 --> 00:38:49,060
What it was associated with
was alternative splicing.

802
00:38:49,060 --> 00:38:52,650
So people who have the variant
associated with lighter skin

803
00:38:52,650 --> 00:38:57,210
have a shorter transcript
that's missing Exon 10,

804
00:38:57,210 --> 00:38:58,220
and it's in frame.

805
00:38:58,220 --> 00:39:00,430
It still codes for a protein,

806
00:39:00,430 --> 00:39:05,040
but it's missing this very
important transmembrane domain.

807
00:39:05,040 --> 00:39:08,113
That's gonna really alter
the function of that protein.

808
00:39:10,270 --> 00:39:12,970
Now, when we look at this gene geneology,

809
00:39:12,970 --> 00:39:16,460
we could see that it has a very old time

810
00:39:16,460 --> 00:39:19,530
to most recent common
ancestor, over 2,000,000 years.

811
00:39:19,530 --> 00:39:22,430
And that's actually a kind of
classic signature of what we

812
00:39:22,430 --> 00:39:24,630
call balancing selection.

813
00:39:24,630 --> 00:39:26,540
So balancing selection,

814
00:39:26,540 --> 00:39:28,170
think of it as sort of the opposite

815
00:39:28,170 --> 00:39:31,910
of recent positive directional selection.

816
00:39:31,910 --> 00:39:34,550
It's maintaining different variation

817
00:39:34,550 --> 00:39:36,630
for a long period of time.

818
00:39:36,630 --> 00:39:39,830
So different variants at this
region have been maintained

819
00:39:39,830 --> 00:39:43,003
for a very long time, millions of years.

820
00:39:44,700 --> 00:39:48,740
So in summary, what was
also interesting is that

821
00:39:48,740 --> 00:39:52,370
for half of these genes,
the ancestral variant,

822
00:39:52,370 --> 00:39:55,190
is the variant associated
with the light skin.

823
00:39:55,190 --> 00:39:59,920
And when we estimate the
age of the derived mutation,

824
00:39:59,920 --> 00:40:04,110
it typically predates the
origin of modern humans.

825
00:40:04,110 --> 00:40:07,990
So that means that both light
and dark associated alleles

826
00:40:07,990 --> 00:40:11,983
have been segregating in
Africa for a very long time.

827
00:40:13,370 --> 00:40:14,360
Now, what does that tell us

828
00:40:14,360 --> 00:40:17,730
about the evolution of skin color?

829
00:40:17,730 --> 00:40:20,983
Well, this poor chimp
lost all its body hair.

830
00:40:21,880 --> 00:40:24,930
We could see that it's
relatively lightly pigmented.

831
00:40:24,930 --> 00:40:27,760
Now I've been told by
anthropologists that some chimps

832
00:40:27,760 --> 00:40:30,250
do have darkly pigmented skin,

833
00:40:30,250 --> 00:40:33,330
but anthropologists have
also speculated that

834
00:40:33,330 --> 00:40:35,750
this might have been the ancestral status,

835
00:40:35,750 --> 00:40:37,053
having lighter skin.

836
00:40:38,180 --> 00:40:40,710
And then when are
ancestors left the forest

837
00:40:40,710 --> 00:40:43,730
and went into the Savannah
there would've been selection

838
00:40:43,730 --> 00:40:47,223
to lose body hair, partly
for thermoregulation.

839
00:40:48,810 --> 00:40:50,940
And when they lost that body hair,

840
00:40:50,940 --> 00:40:54,240
there'd be selection from
more darkly pigmented skin.

841
00:40:54,240 --> 00:40:55,740
Now, whenever you see these drawings,

842
00:40:55,740 --> 00:40:57,950
they always have like really
darkly pigmented people.

843
00:40:57,950 --> 00:40:58,983
But I wonder,

844
00:40:59,920 --> 00:41:03,020
could there actually had been
variation for skin color,

845
00:41:03,020 --> 00:41:06,163
even in Africa over the
past few million years?

846
00:41:07,274 --> 00:41:10,860
And then another really
interesting finding was that

847
00:41:10,860 --> 00:41:13,000
the only other place

848
00:41:13,000 --> 00:41:16,900
where we saw the dark associated alleles,

849
00:41:16,900 --> 00:41:21,240
were in South Asia and Australo-Melanesia.

850
00:41:21,240 --> 00:41:24,860
And those are places where we
see darkly pigmented people.

851
00:41:24,860 --> 00:41:27,280
Anthropologists had
speculated that this might be

852
00:41:27,280 --> 00:41:29,563
due to convergent evolution,

853
00:41:30,730 --> 00:41:33,980
adaptation to being in
a high UV environment.

854
00:41:33,980 --> 00:41:36,760
But we showed that actually
these were introduced,

855
00:41:36,760 --> 00:41:39,930
they're on the same haplotype
background as in Africa.

856
00:41:39,930 --> 00:41:43,457
So they were introduced during
the out of Africa migration

857
00:41:43,457 --> 00:41:47,233
and then they were maintained
due to natural selection.

858
00:41:49,620 --> 00:41:53,430
So in conclusion, people often ask me,

859
00:41:53,430 --> 00:41:54,890
are and humans still evolving?

860
00:41:54,890 --> 00:41:57,080
And I would say absolutely, yes.

861
00:41:57,080 --> 00:41:59,780
We've seen an example,
certainly at lactose tolerance

862
00:41:59,780 --> 00:42:02,793
where there has been very
rapid recent evolution.

863
00:42:04,010 --> 00:42:06,300
And with that, I would
just think that many people

864
00:42:06,300 --> 00:42:08,700
who've contributed to these projects

865
00:42:08,700 --> 00:42:11,340
and my funding organizations.

866
00:42:11,340 --> 00:42:13,590
And with that, I'm happy
to answer any questions

867
00:42:13,590 --> 00:42:14,533
that you may have.

868
00:42:19,130 --> 00:42:20,690
And I'll leave the slides up for a bit

869
00:42:20,690 --> 00:42:23,323
just in case I need to refer to anything.

870
00:42:29,860 --> 00:42:32,570
- Good evening. This is Dr. Ethegwood,

871
00:42:32,570 --> 00:42:34,150
from biology department.

872
00:42:34,150 --> 00:42:36,873
I'm going to ask the first question.

873
00:42:37,810 --> 00:42:40,777
Wonderful talk, Dr. Tishkoff.

874
00:42:41,780 --> 00:42:45,770
I'm curious to know if you
are conducting any studies

875
00:42:45,770 --> 00:42:50,770
on how mitochondrial haplotype
affects phenotypic variation,

876
00:42:52,250 --> 00:42:56,943
as seen on low haploble variations.

877
00:42:58,310 --> 00:43:00,000
- Well, that's a really good question.

878
00:43:00,000 --> 00:43:03,200
So we have some mitochondrial
genome variations,

879
00:43:03,200 --> 00:43:04,530
so that's the genome inherited

880
00:43:04,530 --> 00:43:06,760
in the mitochondria organelle.

881
00:43:06,760 --> 00:43:11,370
It's circular, small, it's
only 16,000 or so nucleotides.

882
00:43:11,370 --> 00:43:13,750
We can actually sequence the whole genome,

883
00:43:13,750 --> 00:43:16,640
but it's only passed on
through the maternal lineage.

884
00:43:16,640 --> 00:43:19,890
And so it can be an interesting
marker for tracing ancestry.

885
00:43:19,890 --> 00:43:20,970
And so we have done that.

886
00:43:20,970 --> 00:43:23,380
We've used it to look at
some of the migrations

887
00:43:23,380 --> 00:43:26,120
and population history in Africa.

888
00:43:26,120 --> 00:43:28,950
But we haven't yet looked
at the impact on phenotype.

889
00:43:28,950 --> 00:43:31,230
I think that would be really interesting.

890
00:43:31,230 --> 00:43:34,000
There's actually a guy named Doug Wallace,

891
00:43:34,000 --> 00:43:39,000
professor at children's
hospital, which neighbors us.

892
00:43:40,630 --> 00:43:43,740
And he's been doing re search
on that for very long time.

893
00:43:43,740 --> 00:43:45,850
And I actually think it
would be really interesting

894
00:43:45,850 --> 00:43:47,033
to take a look at that.

895
00:43:48,000 --> 00:43:48,833
- Thank you.

896
00:43:50,090 --> 00:43:52,600
- Dr.Trishkoff, wonderful talk.

897
00:43:52,600 --> 00:43:55,850
I have another question
from the audience for you,

898
00:43:55,850 --> 00:43:57,140
and it reads as follows,

899
00:43:57,140 --> 00:43:59,610
since more and more
people are tending to have

900
00:43:59,610 --> 00:44:02,900
and develop lactose intolerance,
would it be possible

901
00:44:02,900 --> 00:44:05,923
that eventually all of the
human population could have it?

902
00:44:08,268 --> 00:44:11,006
- Wait, did you say
evolving lactose tolerance?

903
00:44:11,006 --> 00:44:14,101
- Yep, that's what the question is.

904
00:44:14,101 --> 00:44:15,570
- It's certainly possible.

905
00:44:15,570 --> 00:44:19,470
I mean, one thing is that
this variant can be introduced

906
00:44:19,470 --> 00:44:23,340
through interbreeding,
right, we call it admixture.

907
00:44:23,340 --> 00:44:27,950
And so since we're becoming
more and more of an urban,

908
00:44:27,950 --> 00:44:29,590
there are so many urban places

909
00:44:29,590 --> 00:44:32,130
that urbanization is
occurring around the world.

910
00:44:32,130 --> 00:44:33,930
And with that comes a lot of admixture.

911
00:44:33,930 --> 00:44:36,730
So you could imagine it
gets introduced that way,

912
00:44:36,730 --> 00:44:40,040
but it takes a bit of time for
natural selection to occur.

913
00:44:40,040 --> 00:44:43,750
So typically it would only
become common if that person is,

914
00:44:43,750 --> 00:44:47,090
if the population is
drinking a lot of milk.

915
00:44:47,090 --> 00:44:48,920
So it also tends to
co-evolve with culture,

916
00:44:48,920 --> 00:44:51,360
but there are some
culture of modifications

917
00:44:51,360 --> 00:44:52,730
that people have made,

918
00:44:52,730 --> 00:44:53,840
so that they don't need this.

919
00:44:53,840 --> 00:44:57,220
So for example, if you make yogurt,

920
00:44:57,220 --> 00:45:00,300
if you ferment milk and you make cheese,

921
00:45:00,300 --> 00:45:02,290
then the lactose gets broken down.

922
00:45:02,290 --> 00:45:03,630
So you don't actually need that.

923
00:45:03,630 --> 00:45:06,110
So in the Middle East, for
example, they've gotten around,

924
00:45:06,110 --> 00:45:08,220
they don't have to have
this genetic mutation.

925
00:45:08,220 --> 00:45:09,730
I find it fascinating.

926
00:45:09,730 --> 00:45:12,620
Why do Europeans, why Africans have those,

927
00:45:12,620 --> 00:45:15,393
not have the cultural adaptations?

928
00:45:19,390 --> 00:45:23,010
- Let me ask another
question from the audience.

929
00:45:23,010 --> 00:45:26,630
This one says, does darker
skin in the equatorial region

930
00:45:27,550 --> 00:45:30,793
confer protection against
sun damage to DNA?

931
00:45:31,920 --> 00:45:33,033
- Absolutely.

932
00:45:35,090 --> 00:45:37,990
I could tell you that one of the regions

933
00:45:37,990 --> 00:45:41,470
where we do research is in Cameroon.

934
00:45:41,470 --> 00:45:43,050
And when I went to Cameroon,

935
00:45:43,050 --> 00:45:46,690
I was just amazed by how
many albinos I saw there.

936
00:45:46,690 --> 00:45:48,960
It's actually something I'm
really interested in studying.

937
00:45:48,960 --> 00:45:52,660
What is the cause of albinism
being so common in Cameroon?

938
00:45:52,660 --> 00:45:55,465
Like every street corner
I was seeing albinos.

939
00:45:55,465 --> 00:45:57,920
And like, you're just
not used to seeing that.

940
00:45:57,920 --> 00:46:01,600
Now, I've also met with a
number of representatives

941
00:46:01,600 --> 00:46:04,330
of some of the albino
organizations and asked them

942
00:46:04,330 --> 00:46:07,230
what some of the problems
are that they're having,

943
00:46:07,230 --> 00:46:08,183
there's a lot of problems,

944
00:46:08,183 --> 00:46:10,720
so one is there's a lot of stigmatization

945
00:46:10,720 --> 00:46:13,360
and traditionally they're
actually killed at birth

946
00:46:13,360 --> 00:46:15,610
if somebody has an albino baby.

947
00:46:15,610 --> 00:46:18,580
Now in Cameroon they're
becoming more and more accepted.

948
00:46:18,580 --> 00:46:20,020
There's actually a chief now

949
00:46:20,020 --> 00:46:22,550
of one of the villages who's albino.

950
00:46:22,550 --> 00:46:25,410
In Tanzania people are maimed and killed

951
00:46:25,410 --> 00:46:26,830
because they believe that

952
00:46:26,830 --> 00:46:28,660
body parts of albinos bring good luck.

953
00:46:28,660 --> 00:46:30,430
It's really horrible.

954
00:46:30,430 --> 00:46:32,930
And I actually hope that
maybe through education,

955
00:46:32,930 --> 00:46:34,220
if we could explain to people

956
00:46:34,220 --> 00:46:36,550
that there's just one
little change in the genome

957
00:46:36,550 --> 00:46:39,440
that causes this and
they're just like them

958
00:46:39,440 --> 00:46:40,500
in every other respect.

959
00:46:40,500 --> 00:46:41,530
But the people that I,

960
00:46:41,530 --> 00:46:43,810
they also have a lot of health issues,

961
00:46:43,810 --> 00:46:46,773
and for example, they get
skin cancer in their teens.

962
00:46:49,041 --> 00:46:51,000
Because there's not pigment in their eyes,

963
00:46:51,000 --> 00:46:53,550
they get really bad sun damage in eyes.

964
00:46:53,550 --> 00:46:55,760
They lose their vision early,

965
00:46:55,760 --> 00:46:57,110
often they can't even make it

966
00:46:57,110 --> 00:46:58,750
past eighth grade because of that.

967
00:46:58,750 --> 00:47:01,910
So there are some really
serious health consequences

968
00:47:01,910 --> 00:47:04,093
if you don't have that kind of protection.

969
00:47:05,910 --> 00:47:08,270
- Dr. Trishkoff, another question for you,

970
00:47:08,270 --> 00:47:10,330
actually from one of my students,

971
00:47:10,330 --> 00:47:11,340
and the question reads,

972
00:47:11,340 --> 00:47:13,110
do you expect your findings

973
00:47:13,110 --> 00:47:15,290
will eventually alter overall healthcare

974
00:47:15,290 --> 00:47:17,823
when it comes to African
Americans and Africans?

975
00:47:19,650 --> 00:47:20,900
- We hope so.

976
00:47:20,900 --> 00:47:23,550
I didn't talk quite so much
today about some of the studies

977
00:47:23,550 --> 00:47:25,620
that we're doing, looking at, for example,

978
00:47:25,620 --> 00:47:28,243
the genetics of cardiometabolic traits.

979
00:47:29,210 --> 00:47:32,270
Also looking at genetic variation at genes

980
00:47:32,270 --> 00:47:34,940
that play a role in drug metabolism,

981
00:47:34,940 --> 00:47:37,740
which may be very likely or adaptive.

982
00:47:37,740 --> 00:47:39,450
And so you see different frequencies

983
00:47:39,450 --> 00:47:40,610
in different populations.

984
00:47:40,610 --> 00:47:44,010
So people may actually
respond differently to drugs.

985
00:47:44,010 --> 00:47:47,010
And so we do think it's a
way to promote health equity,

986
00:47:47,010 --> 00:47:50,220
to promote studies of more
ethnically diverse populations.

987
00:47:50,220 --> 00:47:52,577
I actually just started a
center for global genomics

988
00:47:52,577 --> 00:47:54,310
and health equity at Penn.

989
00:47:54,310 --> 00:47:57,090
And that's actually one of our
goals, is to increase studies

990
00:47:57,090 --> 00:47:59,160
of more diverse populations.

991
00:47:59,160 --> 00:48:00,230
Because if we don't,

992
00:48:00,230 --> 00:48:04,650
they're not gonna benefit
from the genomics revolution.

993
00:48:04,650 --> 00:48:07,140
We now know that a number
of methods that are used

994
00:48:07,140 --> 00:48:09,230
to predict risk for disease, for example,

995
00:48:09,230 --> 00:48:10,710
have been developed in Europeans,

996
00:48:10,710 --> 00:48:12,930
because that's where all
the studies are done.

997
00:48:12,930 --> 00:48:15,200
But they do terribly
when you try to predict

998
00:48:15,200 --> 00:48:17,210
risk for disease and other ancestries.

999
00:48:17,210 --> 00:48:18,070
That's just horrible.

1000
00:48:18,070 --> 00:48:20,530
So we really to promote equity,

1001
00:48:20,530 --> 00:48:22,260
I think we need to promote studies

1002
00:48:22,260 --> 00:48:25,270
of more diverse populations.

1003
00:48:25,270 --> 00:48:26,103
- Thank you.

1004
00:48:28,310 --> 00:48:31,370
- I'll ask another
question from our audience.

1005
00:48:31,370 --> 00:48:32,510
Could you expand more

1006
00:48:32,510 --> 00:48:36,290
on linkage disequilibrium
with skin colors,

1007
00:48:36,290 --> 00:48:40,123
specifically with your
own research and findings?

1008
00:48:42,660 --> 00:48:45,290
- Okay, that's a tough one. (laughs)

1009
00:48:45,290 --> 00:48:48,216
That's a detailed genetic question there.

1010
00:48:48,216 --> 00:48:53,216
I'm not sure the context in
which I might have said that,

1011
00:48:54,160 --> 00:48:59,160
but, so as an example,
let's go to this one,

1012
00:48:59,720 --> 00:49:01,243
actually, I think DDB1.

1013
00:49:03,580 --> 00:49:06,870
So it would be very similar
to what we were talking about

1014
00:49:06,870 --> 00:49:10,660
with that pitch of the red
extended haplotypes, right?

1015
00:49:10,660 --> 00:49:15,100
That the non-random
association of variants

1016
00:49:15,100 --> 00:49:18,860
is referred to as linkage disequilibrium,

1017
00:49:18,860 --> 00:49:20,470
and a number of things can cause it,

1018
00:49:20,470 --> 00:49:22,580
but one of them is natural selection,

1019
00:49:22,580 --> 00:49:23,880
that when you have this new mutation

1020
00:49:23,880 --> 00:49:25,840
and it rises to high frequency,

1021
00:49:25,840 --> 00:49:28,083
all the variants nearby are linked to it.

1022
00:49:29,060 --> 00:49:32,470
So, we actually see a very similar signal.

1023
00:49:32,470 --> 00:49:34,890
There's sort of other
tests that you can use.

1024
00:49:34,890 --> 00:49:37,830
And we see that at this region actually.

1025
00:49:37,830 --> 00:49:40,540
So there is evidence of strong,

1026
00:49:40,540 --> 00:49:43,320
positive selection in non-Africans

1027
00:49:43,320 --> 00:49:46,240
and particularly in Asia,

1028
00:49:46,240 --> 00:49:49,133
like really, really strong
signature selection.

1029
00:49:50,121 --> 00:49:51,040
And we don't know why.

1030
00:49:51,040 --> 00:49:52,530
I think that's an interesting question.

1031
00:49:52,530 --> 00:49:54,900
I'm not so sure it has
to do with skin color.

1032
00:49:54,900 --> 00:49:56,800
These genes are pleiotropic

1033
00:49:56,800 --> 00:50:00,223
meaning that they can affect
many different traits.

1034
00:50:03,550 --> 00:50:08,550
- Dr. Tishcoff, two questions
that I'll try to combine.

1035
00:50:09,840 --> 00:50:11,820
I think the first part is a sort of

1036
00:50:11,820 --> 00:50:15,023
clarifying broad-brush question.

1037
00:50:16,080 --> 00:50:19,150
Why are genetic populations in Africa,

1038
00:50:19,150 --> 00:50:22,363
so broadly, are varied,

1039
00:50:23,880 --> 00:50:24,960
and in your,

1040
00:50:24,960 --> 00:50:29,170
you did mention the work that
you did in the United States.

1041
00:50:29,170 --> 00:50:32,830
And so the second part of the question is,

1042
00:50:32,830 --> 00:50:36,170
conclusions you may have
drawn from genetic variation

1043
00:50:36,170 --> 00:50:41,170
in African Africans
versus African Americans.

1044
00:50:43,469 --> 00:50:45,810
- So first question in terms of

1045
00:50:45,810 --> 00:50:47,860
what's causing all that diversity?

1046
00:50:47,860 --> 00:50:50,253
So let's go back, let's go to this slide,

1047
00:50:52,700 --> 00:50:56,033
that really, I think nicely
illustrates all that diversity.

1048
00:50:57,010 --> 00:51:01,290
So, what's causing it
is the fact that what,

1049
00:51:01,290 --> 00:51:04,440
they arose 300,000 years ago.

1050
00:51:04,440 --> 00:51:07,810
So there have been modern humans living

1051
00:51:07,810 --> 00:51:10,370
on the continent of Africa
longer than anywhere else, right?

1052
00:51:10,370 --> 00:51:12,810
So you have a long period of time.

1053
00:51:12,810 --> 00:51:15,570
Two is that you've got people who migrate

1054
00:51:15,570 --> 00:51:17,280
to different regions.

1055
00:51:17,280 --> 00:51:18,970
They may be separated from each other

1056
00:51:18,970 --> 00:51:20,220
for long periods of time,

1057
00:51:20,220 --> 00:51:21,940
could be due to changes in climate.

1058
00:51:21,940 --> 00:51:24,683
So things like the Saharan
desert is a major barrier,

1059
00:51:25,710 --> 00:51:28,240
and the climate really shifts over time.

1060
00:51:28,240 --> 00:51:31,100
There could be cultural
difference and so on.

1061
00:51:31,100 --> 00:51:34,180
And so they remain separate populations

1062
00:51:34,180 --> 00:51:37,507
and they're sort of
accumulating differences

1063
00:51:37,507 --> 00:51:38,770
and frequency of variants,

1064
00:51:38,770 --> 00:51:41,200
either due to what we
call random genetic drift,

1065
00:51:41,200 --> 00:51:45,250
just random fluctuations or
due to natural selection.

1066
00:51:45,250 --> 00:51:47,240
And then you've got on top of that though,

1067
00:51:47,240 --> 00:51:48,290
all this migration.

1068
00:51:48,290 --> 00:51:50,850
So like in orange, this is
the migration of the Bantu.

1069
00:51:50,850 --> 00:51:53,627
This is representing that
West African ancestry.

1070
00:51:53,627 --> 00:51:56,170
And you could see how
they migrated to the east,

1071
00:51:56,170 --> 00:51:58,200
into the south and then admix,

1072
00:51:58,200 --> 00:52:00,510
and then the question
about African Americans.

1073
00:52:00,510 --> 00:52:02,730
Truthfully, we're not right now focused

1074
00:52:02,730 --> 00:52:05,130
on studying African Americans.

1075
00:52:05,130 --> 00:52:06,660
That was a collaboration.

1076
00:52:06,660 --> 00:52:09,140
We looked at variation
in African Americans,

1077
00:52:09,140 --> 00:52:11,880
in collaboration with some other groups.

1078
00:52:11,880 --> 00:52:13,280
There are other people

1079
00:52:13,280 --> 00:52:15,440
who are doing studies
in African Americans.

1080
00:52:15,440 --> 00:52:16,610
There's a lot of interest

1081
00:52:16,610 --> 00:52:19,940
in trying to infer ancestry from Africa.

1082
00:52:19,940 --> 00:52:21,830
That's very complex.

1083
00:52:21,830 --> 00:52:24,910
It has not, in my opinion,
been done well yet.

1084
00:52:24,910 --> 00:52:28,620
And that's because most of
the African American ancestry

1085
00:52:28,620 --> 00:52:31,250
is from these groups
that are shown in orange.

1086
00:52:31,250 --> 00:52:34,703
And they're pretty genetically
homogeneous, actually,

1087
00:52:35,690 --> 00:52:37,570
if you look at these guys
compared to each other,

1088
00:52:37,570 --> 00:52:39,980
look at these four populations, right,

1089
00:52:39,980 --> 00:52:41,410
and compare them to these.

1090
00:52:41,410 --> 00:52:42,480
These guys are,

1091
00:52:42,480 --> 00:52:45,150
these populations are pretty homogeneous.

1092
00:52:45,150 --> 00:52:47,490
And so it's really hard to know, well,

1093
00:52:47,490 --> 00:52:50,780
which population did
the ancestry come from?

1094
00:52:50,780 --> 00:52:52,550
And the truth is that over time,

1095
00:52:52,550 --> 00:52:55,100
because of interbreeding
between different people

1096
00:52:55,100 --> 00:52:56,730
from different parts of Africa,

1097
00:52:56,730 --> 00:52:58,750
you may see that parts of the genome,

1098
00:52:58,750 --> 00:53:00,600
one part might be from this region,

1099
00:53:00,600 --> 00:53:02,050
another part might be from this region,

1100
00:53:02,050 --> 00:53:03,700
another part from this region,

1101
00:53:03,700 --> 00:53:05,260
and then as I said, we
don't have anything,

1102
00:53:05,260 --> 00:53:06,750
very little from Angola.

1103
00:53:06,750 --> 00:53:08,403
So we're missing a lot of that,

1104
00:53:09,290 --> 00:53:10,910
but also it is really important

1105
00:53:10,910 --> 00:53:12,150
there are health implications,

1106
00:53:12,150 --> 00:53:15,820
that it's really important
to understand both genetic,

1107
00:53:15,820 --> 00:53:19,700
but also environmental social
risk factors for disease

1108
00:53:19,700 --> 00:53:22,993
in ethnically diverse populations
to promote health equity.

1109
00:53:24,010 --> 00:53:24,843
- Thank you.

1110
00:53:26,210 --> 00:53:29,100
- I'll ask another
question from the audience.

1111
00:53:29,100 --> 00:53:31,990
How do you explain that Mbuti pygmy

1112
00:53:31,990 --> 00:53:36,140
speak a Nilo-Saharan
language or Central Sudan

1113
00:53:37,020 --> 00:53:41,240
and that the Luo of Kenya
are genetically Bantu

1114
00:53:41,240 --> 00:53:46,240
or West African, yet also
speak Nilo-Saharan language?

1115
00:53:48,780 --> 00:53:50,240
- What was the part about the Kenyan?

1116
00:53:50,240 --> 00:53:51,290
Say it one more time.

1117
00:53:52,840 --> 00:53:55,060
- So how do you explain the Mbuti pygmy,

1118
00:53:55,060 --> 00:53:56,720
- Yeah, that one with the Nilo-Saharan

1119
00:53:56,720 --> 00:53:58,050
and then what was the next part?

1120
00:53:58,050 --> 00:54:01,314
- And the next part was, and that the.

1121
00:54:01,314 --> 00:54:02,710
- Luo?

1122
00:54:02,710 --> 00:54:03,660
- Luo of Kenya,

1123
00:54:03,660 --> 00:54:06,083
- Okay, that's what I thought,
I just wanted to make sure

1124
00:54:06,083 --> 00:54:08,740
- Are Bantu and also speak Nilo-Saharan?

1125
00:54:09,600 --> 00:54:11,360
- Okay, so let's look at the Mbuti.

1126
00:54:11,360 --> 00:54:14,000
Let's see, here are the Mbuti right here.

1127
00:54:14,000 --> 00:54:18,490
And they're a hunting-gathering population

1128
00:54:18,490 --> 00:54:19,940
they're from Central Africa.

1129
00:54:19,940 --> 00:54:21,480
One of the interesting things about them,

1130
00:54:21,480 --> 00:54:24,320
is this looks like shared
ancestry with the San.

1131
00:54:24,320 --> 00:54:26,190
I find that fascinating.

1132
00:54:26,190 --> 00:54:28,710
Now all so-called pygmy populations,

1133
00:54:28,710 --> 00:54:30,710
these Central African hunter-gatherers,

1134
00:54:30,710 --> 00:54:33,430
now speak the languages
of neighboring groups.

1135
00:54:33,430 --> 00:54:36,730
And they have been kind of dominated.

1136
00:54:36,730 --> 00:54:38,250
It's a really like a cast system.

1137
00:54:38,250 --> 00:54:39,430
It's really horrible.

1138
00:54:39,430 --> 00:54:41,760
I mean, I have actually spoken

1139
00:54:41,760 --> 00:54:44,180
to some of the Bantu speaking populations.

1140
00:54:44,180 --> 00:54:45,870
They say they own the Pygmies,

1141
00:54:45,870 --> 00:54:47,640
I mean, they really believe this,

1142
00:54:47,640 --> 00:54:50,270
it's shocking that this happens today,

1143
00:54:50,270 --> 00:54:52,230
but this has actually
probably been going on

1144
00:54:52,230 --> 00:54:54,190
for a number of thousands of years.

1145
00:54:54,190 --> 00:54:57,090
So the Pygmies have adopted the language

1146
00:54:57,090 --> 00:54:59,530
of whatever the neighboring groups are.

1147
00:54:59,530 --> 00:55:02,270
And my understanding, if you
look at where the Mbuti are,

1148
00:55:02,270 --> 00:55:06,600
look at this, are really close
to where Nilo-Saharans are.

1149
00:55:06,600 --> 00:55:09,470
Now, I don't see a huge amount
of Nilo-Saharan ancestry.

1150
00:55:09,470 --> 00:55:11,740
The Nilo-Saharan ancestry is shown in red.

1151
00:55:11,740 --> 00:55:13,990
I've seen a little bit,
I'm not seeing a lot,

1152
00:55:13,990 --> 00:55:15,980
but it could just be that
they adopted the language

1153
00:55:15,980 --> 00:55:17,010
of the neighboring people.

1154
00:55:17,010 --> 00:55:18,270
Now at the same time,

1155
00:55:18,270 --> 00:55:20,990
I will say that some
recent research we're doing

1156
00:55:20,990 --> 00:55:25,200
showing some very interesting
connections with Nilo-Saharans

1157
00:55:25,200 --> 00:55:28,100
that there might have been
some ancient gene flow.

1158
00:55:28,100 --> 00:55:29,840
And we're still trying to figure that out.

1159
00:55:29,840 --> 00:55:31,070
Now, the Luo.

1160
00:55:31,070 --> 00:55:33,300
All right, let's look at the Luo.

1161
00:55:33,300 --> 00:55:34,720
It's kind of cool, Oh God!

1162
00:55:34,720 --> 00:55:37,080
I'm never gonna be able to read this.

1163
00:55:37,080 --> 00:55:38,730
I know. (laughs)

1164
00:55:38,730 --> 00:55:39,563
Here they are.

1165
00:55:39,563 --> 00:55:40,893
I actually saw, okay.

1166
00:55:42,480 --> 00:55:44,460
Here are the Luo, all right.

1167
00:55:44,460 --> 00:55:47,100
So the Luo, that you could see by the red,

1168
00:55:47,100 --> 00:55:48,980
they speak Nilo-Saharan language,

1169
00:55:48,980 --> 00:55:50,620
but look where they're clustering.

1170
00:55:50,620 --> 00:55:53,510
They're right here with
all the Bantu speakers.

1171
00:55:53,510 --> 00:55:55,090
That is a great example

1172
00:55:55,090 --> 00:56:00,090
where there has been so much admixture

1173
00:56:00,650 --> 00:56:03,760
that there's been almost a
turnover in the genetic variation

1174
00:56:03,760 --> 00:56:06,870
it's largely Bantu, but they
maintain their language,

1175
00:56:06,870 --> 00:56:07,970
their ancestral language.

1176
00:56:07,970 --> 00:56:09,490
And that's a really interesting question

1177
00:56:09,490 --> 00:56:11,650
from an anthropological perspective.

1178
00:56:11,650 --> 00:56:15,240
Why do some populations
maintain their language

1179
00:56:15,240 --> 00:56:18,150
in the face of all this
admixtures and others don't?

1180
00:56:18,150 --> 00:56:22,080
So actually that's the
tribe that Barrack Obama,

1181
00:56:22,080 --> 00:56:24,113
his father was from the Luo tribe.

1182
00:56:25,810 --> 00:56:27,710
So anyhow, that's the reason for that.

1183
00:56:28,770 --> 00:56:29,950
- Thank you.

1184
00:56:29,950 --> 00:56:31,267
- Dr. Trishkoff ,I think we've got time

1185
00:56:31,267 --> 00:56:32,660
for two more questions.

1186
00:56:32,660 --> 00:56:35,340
So I'll take a flier at one.

1187
00:56:35,340 --> 00:56:37,410
Do you have any potential hypothesis

1188
00:56:37,410 --> 00:56:40,987
for why there are so many
albino individuals in cameroon?

1189
00:56:40,987 --> 00:56:43,300
- Oh, I got all kinds of hypothesis.

1190
00:56:43,300 --> 00:56:44,910
No proof. (laughs)

1191
00:56:44,910 --> 00:56:46,510
I have, honestly, I don't know.

1192
00:56:46,510 --> 00:56:49,960
I mean, the theme that I
found intriguing is the fact

1193
00:56:49,960 --> 00:56:54,400
that there's such obvious
selection, negative selection,

1194
00:56:54,400 --> 00:56:56,270
against being albino in Africa, right?

1195
00:56:56,270 --> 00:56:58,630
You're not, you know,
without any intervention,

1196
00:56:58,630 --> 00:57:00,610
you're gonna do pretty poorly.

1197
00:57:00,610 --> 00:57:02,760
Health wise, you're stigmatized.

1198
00:57:02,760 --> 00:57:04,620
They're often killed at birth.

1199
00:57:04,620 --> 00:57:06,520
This is about a strong negative selection

1200
00:57:06,520 --> 00:57:07,900
as you could possibly have.

1201
00:57:07,900 --> 00:57:11,460
So why does this variation maintain?

1202
00:57:11,460 --> 00:57:12,580
Some people have argued

1203
00:57:12,580 --> 00:57:14,410
that it was what's called a founder event

1204
00:57:14,410 --> 00:57:18,300
so that it might have originated
in certain Bantu groups

1205
00:57:18,300 --> 00:57:21,870
that the founder of those
populations had this variant

1206
00:57:21,870 --> 00:57:25,210
and sometimes in small
populations through random drift,

1207
00:57:25,210 --> 00:57:27,550
you can get variance shift in frequency.

1208
00:57:27,550 --> 00:57:29,390
So that's one theory,

1209
00:57:29,390 --> 00:57:30,610
but the one that I have,

1210
00:57:30,610 --> 00:57:33,600
is I think there's some kind
of heterozygote advantage.

1211
00:57:33,600 --> 00:57:35,760
And that's why when we
go back to Cameroon,

1212
00:57:35,760 --> 00:57:37,500
we want to study that,

1213
00:57:37,500 --> 00:57:39,653
and try to determine what that might be.

1214
00:57:40,790 --> 00:57:41,623
- Right.

1215
00:57:42,650 --> 00:57:44,933
Dr. Gordon, do you wanna
ask the final question?

1216
00:57:44,933 --> 00:57:47,087
- I'm gonna ask this one.

1217
00:57:47,087 --> 00:57:50,740
Do you think you will ever be able to go

1218
00:57:50,740 --> 00:57:53,900
or to plan to travel
to Angola in the future

1219
00:57:53,900 --> 00:57:56,970
and to study the future
into the slave trade

1220
00:57:56,970 --> 00:58:01,680
and how that relates to the
genetic diversity and evolution?

1221
00:58:01,680 --> 00:58:03,930
- Well, it would certainly
be great. (chuckles)

1222
00:58:03,930 --> 00:58:06,700
I find that there's just so
many interesting places to go,

1223
00:58:06,700 --> 00:58:07,780
not enough time.

1224
00:58:07,780 --> 00:58:11,190
And it partly depends on having partners,

1225
00:58:11,190 --> 00:58:14,010
and the safety, and the politics.

1226
00:58:14,010 --> 00:58:15,260
I do work in Ethiopia.

1227
00:58:15,260 --> 00:58:17,940
I can't go to Ethiopia now
because of the politics.

1228
00:58:17,940 --> 00:58:19,890
I can't go back to Northwestern Cameroon

1229
00:58:19,890 --> 00:58:22,290
because of the politics,
so we'll have to see,

1230
00:58:22,290 --> 00:58:24,620
but one group that we have studied,

1231
00:58:24,620 --> 00:58:27,600
we have done research in Botswana,

1232
00:58:27,600 --> 00:58:30,070
and there's a group in
the Northwest of Botswana

1233
00:58:30,070 --> 00:58:32,360
that came from Angola about 100 years ago,

1234
00:58:32,360 --> 00:58:35,883
within the past 100 years as refugees.

1235
00:58:38,010 --> 00:58:41,910
So by studying them, we might
get a little bit of a clue,

1236
00:58:41,910 --> 00:58:44,910
but I think there are other
groups, other researchers

1237
00:58:44,910 --> 00:58:46,510
who are also interested in that.

1238
00:58:47,530 --> 00:58:48,780
- Thank you.

1239
00:58:48,780 --> 00:58:49,760
- Well, Dr. Tishkoff

1240
00:58:49,760 --> 00:58:50,967
thanks for a wonderfully
illuminating talk.

1241
00:58:50,967 --> 00:58:51,800
It is.