Skip to main content
. 2005 May;170(1):291–304. doi: 10.1534/genetics.104.038232

TABLE 3.

Concordance between LD relations and tag SNP selection in five genes for three ethnic populations

Japanese
African Americans
Caucasians
Gene name No. of
SNPs testeda
Average value of
pairwise LDb
Threshold
value of
LD groupingc
No. of tag
SNPs/no. of
polymorphic
SNPsd (%)
Average value of
pairwise LD
Threshold
value of
LD groupingc
No. of tag
SNPs/no. of
polymorphic
SNPsd (%)
Average value of
pairwise LD
Threshold
value of
LD groupingc
No. of tag
SNPs/no. of
polymorphic
SNPsd (%)
ABCA1 37 0.12 0.4 19/37 (51) 0.05 0.2 24/37 (64) 0.08 0.5 17/37 (45)
ADPRT 29 0.38 0.4  7/29 (24) 0.26 0.6 10/28 (35) 0.32 0.2  7/28 (25)
F5 41 0.25 0.7 12/41 (29) 0.18 0.6 15/41 (36) 0.22 0.2 10/40 (25)
LPL 25 0.31 0.5  9/23 (39) 0.14 0.5 17/25 (68) 0.23 0.4 12/25 (48)
SLC12A3 25 0.14 0.3 10/25 (40) 0.05 0.2 19/25 (76) 0.10 0.3 11/25 (44)

ABCA1, ATP-binding cassette, subfamily A (ABC1), member 1; ADPRT, poly (ADP-ribose) polymerase family, member 1; F5, coagulation factor V; SLC12A3, solute carrier family 12 (sodium/chloride transporters), member 3.

a

From exons, 5′-UTRs, and 3′-UTRs of each gene, SNPs with MAF of at least 5% in any of three populations were tested.

b

The LD by coefficient r2 was averaged over all possible pairs of polymorphic SNPs.

c

The optimal r2 threshold for defining LD groups of SNPs was chosen automatically within our algorithm (see the minimum value, t, in step 7).

d

This ratio indicates the efficiency of tag SNP selection (see results).