| Autor: |
Satyawan, Dani, Siregar, Heri Adriwan, Wening, Sri, Lestari, Puji, Suprianto, Edy |
| Předmět: |
|
| Zdroj: |
AIP Conference Proceedings; 2023, Vol. 2972 Issue 1, p1-8, 8p |
| Abstrakt: |
Oil palm is the most productive oil crop in the world but breeding a new oil palm variety is a costly and time-consuming activity due to its long reproductive cycle and large plant size. Genomic selection technology utilizes whole-genome genotypic data to predict the best-performing breeding lines and enable the selection of breeding lines from the seedling stage while maximizing genetic gains in each selection cycle. We explored the utility of genomic selection implemented using the rrBLUP package in predicting the best breeding lines in a population comprising 499 F1, BC1, and parental generations from crosses between various Elaeis guineensis and Elaeis oleifera accessions. Phenotypic scores for 29 traits were collected from each line and combined with genotypic data from 3908 SNP markers to calculate the marker effects for trait predictions. When all lines were treated as a single population, and 60% of the population was selected randomly as the training population, high prediction accuracy was observed in some traits, and 19 traits had prediction accuracy between 0.406 to 0.709. However, the prediction accuracy dropped significantly in most traits when we simulated intergenerational prediction by using only the F1 and parental lines as the training population and used the calculated marker effects to predict the phenotypes of BC1 lines. Only two traits (mesocarp per fruit and leaflet width) retained accuracy values larger than 0.400. Thus, further investigations and optimizations are still required to improve the utility of rrBLUP for intergenerational genomic selection in pseudo-backcross populations. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
