Genomic prediction accuracy in CIMMYT's East African maize breeding program
Material type: TextPublication details: 2014Description: 1 pageSummary: Genomic prediction is a new method that uses markers across the whole genome to predict individual breeding values at an early growth stage. One of the applications of genomic prediction in plant breeding is to help the breeder select beneficial crosses to create the next generation. The aim of this study was to investigate the genomic prediction accuracy of the diverse population of CIMMYT's East African maize breeding program. The genotypic dataset contains 2022 lines and 66K SNP markers from genotyping by sequencing. The phenotypic data consists of yield evaluations from 154 trials. Genomic prediction was done in two steps: first the BLUP for each line was estimated by considering trial, block and tester effects. In the second step direct genomic values (DGVs) were predicted from BLUPs using ridge regression (rrBLUP). Cross validation was used to estimate the accuracies and it was repeated 50 times. In each cross validation a random 30% of the lines were considered the test population and the rest of the lines were the training population. To visualize the diversity of the population principle component analysis (PCA) was done. The results showed there are five clusters of lines in the population. Genomic prediction was done for each cluster and for all lines. The accuracies varied between 0.28 and 0.41 for different clusters and 0.36 for all lines. These accuracies show promise for implementing genomic prediction in CIMMYT's East African maize breeding program.Item type | Current library | Collection | Call number | Status | Date due | Barcode | Item holds | |
---|---|---|---|---|---|---|---|---|
Conference proceedings | CIMMYT Knowledge Center: John Woolston Library | CIMMYT Staff Publications Collection | CIS-7466 (Browse shelf(Opens below)) | Available |
Abstract only
Genomic prediction is a new method that uses markers across the whole genome to predict individual breeding values at an early growth stage. One of the applications of genomic prediction in plant breeding is to help the breeder select beneficial crosses to create the next generation. The aim of this study was to investigate the genomic prediction accuracy of the diverse population of CIMMYT's East African maize breeding program. The genotypic dataset contains 2022 lines and 66K SNP markers from genotyping by sequencing. The phenotypic data consists of yield evaluations from 154 trials. Genomic prediction was done in two steps: first the BLUP for each line was estimated by considering trial, block and tester effects. In the second step direct genomic values (DGVs) were predicted from BLUPs using ridge regression (rrBLUP). Cross validation was used to estimate the accuracies and it was repeated 50 times. In each cross validation a random 30% of the lines were considered the test population and the rest of the lines were the training population. To visualize the diversity of the population principle component analysis (PCA) was done. The results showed there are five clusters of lines in the population. Genomic prediction was done for each cluster and for all lines. The accuracies varied between 0.28 and 0.41 for different clusters and 0.36 for all lines. These accuracies show promise for implementing genomic prediction in CIMMYT's East African maize breeding program.
Global Maize Program|Genetic Resources Program
English
Lucia Segura
INT2765|INT2869|INT3400|CCJL01|INT3333
CIMMYT Staff Publications Collection