Linkage disequilibrium (LD) is a
phenomenon observed in populations where alleles at different loci on a
chromosome are not randomly associated but instead are found together more
frequently or less frequently than expected by chance. In simpler terms, it
refers to the non-random association of alleles at different loci.
LD arises due to several factors, including genetic linkage,
mutation, recombination, genetic drift, and natural selection. When two loci
are in LD, the alleles at these loci are inherited together more often than
expected if they were independent. LD can vary across the genome and between
populations, and it plays a crucial role in various genetic analyses, including
association mapping, evolutionary studies, and population genetics.
Common measures of LD include:
D' (D prime):
·
D'
is a standardized measure of LD that ranges from 0 to 1, with 1 indicating
complete LD, 0 indicating complete equilibrium, and negative values indicating
an excess of recombination.
·
Advantages:
D' is less affected by allele frequencies than other LD measures and is widely
used for comparing LD across loci and populations.
·
Limitations:
D' does not account for differences in allele frequencies between populations
and can be influenced by sample size and marker density.
r^2 (Squared correlation coefficient):
·
r^2
measures the correlation between alleles at two loci and ranges from 0 to 1,
with 1 indicating complete LD and 0 indicating no LD.
·
Advantages:
r^2 provides a straightforward measure of LD strength and is commonly used in
association mapping and haplotype analysis.
·
Limitations:
r^2 is influenced by allele frequencies and may not accurately reflect LD
strength in populations with low minor allele frequencies or rare variants.
Correlation coefficient (Pearson's
correlation coefficient):
·
The
correlation coefficient measures the linear relationship between allele
frequencies at two loci and ranges from -1 to 1, with 1 indicating perfect
positive correlation, -1 indicating perfect negative correlation, and 0
indicating no correlation.
·
Advantages:
The correlation coefficient can provide insights into the direction and
strength of LD between loci.
·
Limitations:
The correlation coefficient may not account for non-linear relationships or
deviations from Hardy-Weinberg equilibrium, and it may be sensitive to outliers
or skewed data.
Normalized mutual information (NMI):
·
NMI
measures the mutual information shared between alleles at two loci, normalized
by the entropy of the allele frequencies.
·
Advantages:
NMI accounts for both linear and non-linear relationships between alleles and
is robust to differences in allele frequencies.
Limitations: NMI may be computationally intensive and less
intuitive to interpret compared to other LD measures.
In summary, linkage disequilibrium (LD) is a non-random
association of alleles at different loci, and it plays a crucial role in genetic
analyses. Common measures of LD, including D', r^2, correlation coefficients,
and normalized mutual information, provide insights into the strength and
structure of LD in populations. Each LD measure has its own advantages and
limitations, and researchers should consider the appropriate measure based on
their study objectives, population characteristics, and data properties.
0 Comments