3 Analysis of Population Structure
51
and jackknife assume near-independence of the markers used, and so the linkage
between nearby markers can be accounted for by performing the bootstrap or
jackknife procedure over larger blocks of the genome or whole chromosomes.
Various software packages that can handle different types of genetic data are able
to calculate many of the distance measures discussed above and give a pairwise
distance matrix as output. The distance matrix can then be used for clustering
using, e.g., a tree construction algorithm. In certain software, both matrix calculation
and tree construction algorithms are available. Some examples of commonly used
software are Mega (Kumar et al. 2016), Arlequin (Excoffier and Lischer 2010), and
Plink (Purcell et al. 2007).
An example tree of West African Yoruba, European French, North African
Mozabites, and Southern African San is given in Fig. 3.1. To construct the tree,
F re n ch H G D P 00 51 1
F r e n c h H G D P 0 0 5 1 2
F
r
e
n
c
h
H
G
D
P
0
0
5
1
3
F r e n c h H G D P 0 0 5 1 4
F re n ch H G D P 00 51 5
Fren ch HGD P00 516
F r e n c h
H G D P 0 0 5 1 7
Fren ch HGD P00 518
F r e n c h H G D P 0 0 5 1 9
F r e n c h
H G
D P 0 0 5 2 0
Y o r u b a H G D P 0 0 9 2 0
Yo ru ba HG DP 00 92 4
Yoruba HGDP00925
Y
o
r
u
b
a
H
G
D
P
0
0
9
2
6
Y o ru b a H G D P 0 0 9 2 7
Y
o
ru
b
a
H
G
D
P
0
0
9
2
8
Y
o
r
u
b
a
H
G
D
P
0
0
9
2
9
Y o r u b a H G D P 0 0 9 3 0
Y o ru b a H G D P 0 0 9 3 1
Yo ru ba HG DP 00 93 2
S a n H G D P 0 0 9 9 1
S a n H G D P 0 0 9 9 2
S a n H G D P 0 1 0 2 9
S a n
H G
D P 0 1 0 3 2
S a n
H G D P 0 1 0 3 6
Mozabit e HGDP0 1253
M oz ab ite H G D P0 12 54
M
o
z
a
b
it
e
H
G
D
P
0
1
2
5
5
M
oz
ab
ite
H
G
D
P0
12
56
M o z a b it e H G D P 0 1 2 5 7
Mo
zab
ite
HG
DP
012
58
Mozabit
e
HGDP0
1259
M o z a b it e H G D P 0 1 2 6 0
Mo zab ite HG DP 012 61
M
o
z
a
b
it
e
H
G
D
P
0
1
2
6
2
Fig. 3.1 An example neighbor-joining tree of individuals from West Africa (Yoruba), Europe
(French), North Africa (Mozabite), and Southern African (San). The NJ tree was built from an
IBS distance matrix (computed in Plink) and the R-package ape. The individuals cluster according
to their sample locations, except for the Mozabites that contain the French sample as a subset
51
and jackknife assume near-independence of the markers used, and so the linkage
between nearby markers can be accounted for by performing the bootstrap or
jackknife procedure over larger blocks of the genome or whole chromosomes.
Various software packages that can handle different types of genetic data are able
to calculate many of the distance measures discussed above and give a pairwise
distance matrix as output. The distance matrix can then be used for clustering
using, e.g., a tree construction algorithm. In certain software, both matrix calculation
and tree construction algorithms are available. Some examples of commonly used
software are Mega (Kumar et al. 2016), Arlequin (Excoffier and Lischer 2010), and
Plink (Purcell et al. 2007).
An example tree of West African Yoruba, European French, North African
Mozabites, and Southern African San is given in Fig. 3.1. To construct the tree,
F re n ch H G D P 00 51 1
F r e n c h H G D P 0 0 5 1 2
F
r
e
n
c
h
H
G
D
P
0
0
5
1
3
F r e n c h H G D P 0 0 5 1 4
F re n ch H G D P 00 51 5
Fren ch HGD P00 516
F r e n c h
H G D P 0 0 5 1 7
Fren ch HGD P00 518
F r e n c h H G D P 0 0 5 1 9
F r e n c h
H G
D P 0 0 5 2 0
Y o r u b a H G D P 0 0 9 2 0
Yo ru ba HG DP 00 92 4
Yoruba HGDP00925
Y
o
r
u
b
a
H
G
D
P
0
0
9
2
6
Y o ru b a H G D P 0 0 9 2 7
Y
o
ru
b
a
H
G
D
P
0
0
9
2
8
Y
o
r
u
b
a
H
G
D
P
0
0
9
2
9
Y o r u b a H G D P 0 0 9 3 0
Y o ru b a H G D P 0 0 9 3 1
Yo ru ba HG DP 00 93 2
S a n H G D P 0 0 9 9 1
S a n H G D P 0 0 9 9 2
S a n H G D P 0 1 0 2 9
S a n
H G
D P 0 1 0 3 2
S a n
H G D P 0 1 0 3 6
Mozabit e HGDP0 1253
M oz ab ite H G D P0 12 54
M
o
z
a
b
it
e
H
G
D
P
0
1
2
5
5
M
oz
ab
ite
H
G
D
P0
12
56
M o z a b it e H G D P 0 1 2 5 7
Mo
zab
ite
HG
DP
012
58
Mozabit
e
HGDP0
1259
M o z a b it e H G D P 0 1 2 6 0
Mo zab ite HG DP 012 61
M
o
z
a
b
it
e
H
G
D
P
0
1
2
6
2
Fig. 3.1 An example neighbor-joining tree of individuals from West Africa (Yoruba), Europe
(French), North Africa (Mozabite), and Southern African (San). The NJ tree was built from an
IBS distance matrix (computed in Plink) and the R-package ape. The individuals cluster according
to their sample locations, except for the Mozabites that contain the French sample as a subset