Based on the phenotypic data, two extreme bulks
(also referred to as pools in this report) were formed in
each population with consideration given to cassava
mosaic disease (CMD) interaction with root yield in
cassava. The consideration was stricter with the
genotypes used to form the LB group. Any CMD
susceptible genotype (CMD severity score of two and
above) with low root yield at 7 MAP was not selected
as a LB genotype (Table 2). DNA bulk representing
each extreme phenotype in a population was formed
by mixing equal volume of DNA solution from each
genotype in the bulk. DNA samples of the two parents
and the two bulks in each population were assayed
using the selected polymorphic markers (pm-p). The
gels were scored for each marker used for the assay as
described above and the polymorphic ones (pm-pb)
were selected for each population. A marker was
regarded to be polymorphic at this stage when the
different bands formed by the EB and LB bulks were
similar to the bands formed by the EB and LB parents,
respectively. Individual genotypes in the contrasting
bulks of each population were then assayed using the
markers found to be polymorphic between the two
bulks (pm-pb) to select the candidate markers.
The genotypic data generated for each marker from
screening of the individuals in the contrasting bulks of
each population were correlated with the phenotypic
data using Pearson correlation test in Microsoft Excel.
Percent of presence or absence of band (EB parent
band) among all the individuals in each phenotypic
group was estimated for each polymorphic marker. A
marker was regarded to be closely associated with a
gene when the correlation coefficient between phenotypic
and genotypic data was C0.3 (Fregene, 2006,