2018 ASHS Annual Conference

Predictive breeding has become more and more popular in efforts to unravelling the genetic control of complex traits in plants. Cowpea (Vigna unguiculata (L.) Walp) is one the most important legumes grown worldwide but significantly affected by various types of abiotic stress. To the best of our knowledge, little has been done with respect to studies pertaining to genomic selection to assist cowpea breeders with efficiently gaining genetic gain. In this study, the focus will be on stresses such as drought, salt, low phosphorus conditions, and iron chlorosis deficiency, which have been proven to substantially limiting cowpea production. Therefore, the objective of this study will be to estimate the accuracy of genomic estimated breeding values (GEBV) for these aforementioned abiotic stresses. Phenotyping on drought, salt, low phosphorus conditions, and iron chlorosis deficiency was conducted on a total of 200, 155, 357, 353 cowpea accessions, respectively. Genomic estimated breeding values will be computed for at least 1000 SNPs postulated from genotyping-by-sequencing. Training set will consist of subsets randomly chosen from the association panel. Set sampling will be conducted using R through 10-fold cross validation at each sampling process. Genomic selection will be performed in R using rrBLUP, Bayes A, Bayes B, and LASSO. We expect that: 1) genomic selection accuracy will vary from low to moderate depending upon the type of abiotic stress, 2) At least 10 SNPs will have large effects for each trait, 3) Bayes B will provide the highest genomic selection accuracy, 4) there will be positive correlation between training/testing set size and genomic selection accuracy. The results from this investigation can be used in breeding cowpea cultivars for more tolerance to drought, salt, low phosphorus conditions, and iron chlorosis deficiency.