Development and testing of a SNP marker platform in sugarcane : final report 2012/025
Abstract
Marker assisted selection is being used in many industries to develop new and improved crops. This
project sought to develop a selection tool for the Australian sugarcane industry by developing a set
of DNA markers that could be used to enhance rates of genetic gain in the Sugarcane Breeding
Program when compared to phenotypic selection alone. This was a collaborative project between
CSIRO, Syngenta and SRA and has been proven successful in combining the skills of each organisation
to achieve the best outcome. The extensive experience within Syngenta on other crops was used in
combination with the sugarcane genetic knowledge within CSIRO to select the Affymetrix Axiom
array system as the platform most likely to generate results on such a complex polyploid as
sugarcane. This is the first time that this technology has been utilised on a species that has a complex
high ploidy genome. To date, genetic markers in sugarcane have lagged behind other crops. This is due to both its
complex genetic structure and large genome size which makes generation of enough markers to
cover the genome very laborious. The array development has removed this obstacle in sugarcane
and made available a tool that can generate thousands of polymorphic markers in a single
experiment. The marker of choice for all plant and animal selection programs has become the single
nucleotide polymorphism (SNP). This is due to their presence throughout the genome in large
numbers, robust detection and applicability to high throughput technologies. The informative markers were used to create a highly informative 40K array, cheaper to screen than
the 400K array. This was used to screen 1850 breeding lines for genomic selection analysis to test the
gain in prediction accuracy from the addition of SNP marker information to the pedigree information.
The results have been highly encouraging and show that SNP information has a much better
predictive accuracy for breeding value than using pedigree information for all traits tested. The
predictive accuracies of around 0.47 for CCS, and 0.3 for TCH, give high confidence that this method
can be used effectively for parental improvement and lead to new varieties with increased CCS and
yield. This project has been very successful with the generation of the highly informative cane array which
has already been shown to generate thousands of SNP markers in the Australian Sugarcane Breeding
Program population. The project organised two workshops on marker development and
implementation in sugarcane which were well attended by both the sugarcane breeders and
molecular geneticists.