The CIAT genotypes, in addition to other important traits which they possess, such as high dry matter content and yield, also served as donor parent for the CMD2 gene in these crosses. The F1 families were therefore expected to show segregation for the CMD2
gene. Although 684 individuals were initially planted,phenotypic data indicated that 525 individuals were resistant with 159 individuals being susceptible. The progenies used in the molecular analysis were those selected for CMD resistance based on phenotypic
scoring and should be expected to express marker alleles or corresponding bands for the CMD2 gene.Rabbi et al. (2014) used a high density single nucleotide polymorphism (SNP) map in a bi-parental mapping population segregating for the dominant monogenic resistance to CMD, and found a single locus with large effect, which is probably the CMD2 locus. In their study they found nearly half of the progeny and the female parent to be resistant to CMD,with the remainder of the F1 individuals showing
disease symptoms ranging from mild to severe. The results of this study indicated that 83 % of the genotypes showed at least one marker allele for the CMD2 gene. The CMD markers identified were around 2–8 cM away from the CMD2 locus, so the markers are not on the gene, which means that recombination could occur, causing a deviation from
a 100 % marker-trait association (Okogbenin et al.2007). Therefore selection is more likely to be accurate if there are multiple markers. One allele linked to the gene is suggestive that the gene is present,but not always. If CMD2 was responsible for all the
resistance observed in the F1 families, it would imply that the markers associated with the gene was 83 % successful in identifying resistant lines in the phenotypically
selected resistant individuals analysed with molecular markers. Generally, CMD2 markers have been found to be 68 % effective in MAS, but this is in populations that were not pre-selected phenotypically.So even though CMD2 is a dominant gene, false positives can occur.