3. Revolutionary science
l
First human genome cost $3,000,000,000
l
My son was sequenced in 2014 in utero for
$500 (7x coverage)
l
You can get a retail 30x (highly quality)
genome today for $1,850 today
l
We now have hundreds of ancient genomes
of humans of various qualities!
l
Half the American population will have high
quality WGS by 2025
4. Three stages of -omics
l
Science as industrial
process (then)
l
Science as a laboratory
process (now)
l
Science as a consumer
good (soon)
5. What is in the world-wide genome?
l
3 billion base pairs
(~3 GB)
l
>80 million SNPs
l
4 million indels
l
~60K structural
variants
1000 Genomes Consortium
6. What is in your genome?
l
3 billion base pairs
(~3 GB)
l
~4 million SNPs
l
~15 million
singletons
l
~500K indels
1000 Genomes Consortium
7. Why is your genome important?
l
You are not your
genome
l
But who you are
begins with your
genome
l
A genome is not
sufficient to make
you, but it is
necessary
8. So why are we focusing on SNPs not
WGS?
l
WGS is expensive
l
It takes longer (more resources)
l
Most of it is not informative
l
We can't interpret a lot of the variation
(e.g., copy number variation)
l
More data takes longer to analyze, even
on powerful computers
9. What's in your SNPs?
l
SNP chip assays survey ~10K to
1 million markers
l
Human chips generally 500K to
1 million markers
l
Only a minority of human SNPs
are captured on a given chip
l
Many of the markers are
intergenic (not among ~20K
genes), while the genic ones
may not be functional
10. A few SNPs go a long way
l
30 “ancestrally informative markers” give
you enough information to tell inter-
continental differences (e.g., African vs.
European)
l
1,000 random SNPs are sufficient for data
heavy population assignment at this scale
l
10,000 random SNPs get you most of the
way to intra-continental differences (e.g.,
China vs. Japan)
l
>100,000 markers lots of power
15. A problem of perception
l
Y and mtDNA lineages are passed on as
a single unit
l
The autosomal genome is a synthesis of
billions of co-inherited blocks with
l
Visualizations of relatedness on the
autosomal genome have to simplify and
reduce numerous distinct genealogies
l
Today the difficulty is representing the
variation, rather than getting it
19. Autosomal ancestry
inference as Reader's Digest
of your genomic history
1)Take thousands of
characters
2)Put them into the
scaffold of a
predetermined plot
3)Reduce down to
fewer than fifty
narrative threads
20. Sometimes the “composition” might
turn out a little confusing, but we
hope it tastes good
1)Take ingredients,
300K SNPs across
thousands of
individuals
2)Create a model
3)Check the
parameters across
the individuals which
are the best fit to the
model
21. Or, made more “sciencey”
1. Take data (e.g., genetic
variation)
2. Take a model (e.g., Hardy-
Weinberg)
3. Plug in values (e.g., 2
populations, 3 populations, 4
populations….)
4. Produce plots!
22. MyOrigins version 2.0, 24
clusters, +6 from 1.0
Ashkenazi
Balkan
East Africa
Finland
Germany
British
Italy
Mbuti Pygmies
North African
North Amerindian
North India
North West Asia
Northeast Asia
Papuan
Slavic
South Amerindian
South India
South West Asia
Sardinian
Southeast Asia
Siberian
Spain
Scandinavian
West Africa
23. Better Native American
clusters
l
North Amerindian
uses Pima Indians +
ftDNA data set
l
Has more Na Dene
signal
l
South Amerindian
uses Amazonian
Indians
l
Has an “Austro-
Melanesian”
signature at low
levels
l
Using this as a
“reference” for all
Amerindians was a
mistake!
24. A Siberian cluster
l
Very few customers
are going to be much
Siberian
l
But Siberians turn
out to be very
important in the
history of Eurasia
and the New World
l
Many Native
Americans will pick
up some of this
signature...and
Europeans
25. A Sardinian cluster
l
Very few customers
are going to be much
Sardinian
l
But Sardinians turn
out to be very
important in the
history of Europe
l
If you have Sardinian
admixture (many
South Europeans
will) it will tell you
about deep roots
26. New Southern European
Clusters
l
The Italian cluster is
biased toward South
Italian samples
l
There has been a fair
amount of gene flow
in Southern Europe,
but a definite “cline”
from Spain to Italy
l
Many Iberians will
have “Italian” and
vice versa
l
The Balkan
cluster defines
the eastern
position of the
Southern
European “cline”
l
It is a composite
of indigenous
groups overlain
with Slavic
migration within
the last ~1,500
years
31. Standing on the shoulders of
two papers
l
Eight thousand
years of natural
selection in Europe,
Mathieson et al.
l
Population genomics
of Bronze Age
Eurasia, Allentoft et
al.
32. Of Gods and Men
l
““There were giants in the
earth in those days; and also
after that, when the sons of
God came in unto the
daughters of men, and they
bare children to them, the
same became mighty men
which were of old, men of
renown.””
35. Collective memory or fiction?
l
Oral memory
degrades
l
Humans tend to
create the same
fictive narratives
l
Legends and myths
have low fidelity with
the reality of the
past
l
Memonic techniques
were much more
advanced in the past
l
The structure of
narrative and its
utility can maintain
its fidelity
36. Ancient Europe was inhabited
by very different people
l
Hunters and
gatherers, the oldest
ones
l
The first farmers
l
And the horse people
l
The genetic distance
between some of
these groups was
comparable to that
between Chinese
and modern North
Europeans
39. It is not indigenous to Europe
●
It is entirely absent
in Mesolithic
Western
Europeans, that is,
the oldest group
●
It is present at very
high frequencies in
both the horse
people and the
farmers
41. It was not common in ancient
Europe
●
Again, absent in
Mesolithic
Western
Europeans
●
Present in
moderately low
frequencies among
farmers and at
moderately high
frequencies among
the horse people
44. It was very common in
ancient Europe
●
All of the old hunters
and gatherers had
the blue eye variety
●
The horse
people and the
farmers did not
have that variant
47. Northern Europeans are new thing
●
The genetic architecture responsible
for the distinctive Northern European
phenotype is very new
●
It almost certainly post-dates the Old
Kingdom of Egypt
●
In other words, this is selection and
evolution occurring during historical
periods!
48. A genetic architecture similar to
Copper Age Europeans
● SLC24A5, fixed “European”
variant
● SLC45A2, heterozygote
● OCA2/HERC2, heterozygote
● KITLG, heterozygote
51. The oldest “Swedes” looked
unique
● Unlike Western hunters and gatherers a group from Motala in
Sweden seem to have many lighter color alleles (though still
darker than modern Northern Europeans)
●
But, they happen to carry a genetic variant that is very rare in
Europe today, that is associated with East Asians and aspects
of their phenotype such as very thick and straight hair
● Motala hunters and gatherers have this gene in frequencies
common in Southern Chinese
52. Behind the mask of myth
●
We now know that Europeans on the edge
of history consisted of very distinct peoples
●
Additionally, the genetic traits we are aware
of suggest they looked very different from
each other, and from one another
Europeans
●
And, we now have evidence in other
societies of folk legends rooted in fact
persisting for thousands of years
53. More than metaphor
The myths and legends of the people of
Europe, and other regions, may harbor
very important and illuminating
anthropological and historical information.
Stay tuned!