Agriculture Reference
In-Depth Information
Table 2.4.
Software used for population structure analysis.
Name
Approach and algorithms
Information needed
Reference
Structure
Bayesian, MCMC
Null loci data
Pritchard et al. 2000
BAPS
Bayesian
Null loci data
www.helsinki.
/bsg/
Structurama
probability distribution,
Bayesian, MCMC
Null loci data
Tess
Bayesian
Null loci and
origin data
Chen et al. 2007
Geneland
Bayesian
Null loci and
origin data
number of clusters within a population and vary depending on the
different parameters used in the analysis (Table 2.4). In order to determine
population structure, various pieces of information are considered, the
most important of which is the number and location of null loci. Geo-
graphical information and a set of possible subpopulations are then
optional inputs depending on the programs.
Bayesian approaches can be used for population structure analysis,
althoughmost software use a frequency-based approach. For that reason,
a lot of information about the individuals used in the study is needed,
such as geographical location, information about possible parents, and
so on. However, with the increasing use of Bayesian approaches, almost
any kind of genotypes, even those with little background information,
can be used to test for population structure.
B. Software Used for Population Association Mapping
There are various software programs that can be used in AM analyses
(Table 2.5). Some of these were reviewed by Zhu et al. (2008) in their
discussion of AM approaches, but we have added various new programs
to the table of software for AM. The programs differ in several respects
depending on the models used, language in which the program is
written, and the analysis that can be achieved by each software.
Marker associations are based on at least two tests: one for the
signi
cance or
LOD
score of the LD between all pairs of markers and
the other on an effect test for each allele at each marker for the character
or phenotypic trait(s) under study. To be signi
first test requires
an
LOD
score higher than the threshold, while the second test requires
large effects of one of the alleles. When both of the tests have meaningful
values, then an association for an allele at a locus is found.
cant, the
