Skip to main content
ARS Home » Midwest Area » Madison, Wisconsin » Vegetable Crops Research » Research » Publications at this Location » Publication #336446

Research Project: Cranberry Genetic Improvement and Insect Pest Management

Location: Vegetable Crops Research

Title: Multivariate GBLUP Improves Accuracy of Genomic Selection for Yield and Fruit Weight in Biparental Populations of Vaccinium macrocarpon Ait

Author
item COVARRUBIAS-PAZARAN, GIOVANNY - University Of Wisconsin
item SCHLAUTMAN, BRANDON - University Of Wisconsin
item DIAZ-GARCIA, LUIS - University Of Wisconsin
item GRYGLESKI, EDWARD - Valley Corporation
item Polashock, James
item JOHNSON-CICALESE, JENNIFER - Rutgers University
item VORSA, NICHOLI - Rutgers University
item IORIZZO, MASSIMO - North Carolina State University
item Zalapa, Juan

Submitted to: Frontiers in Plant Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/20/2018
Publication Date: 9/12/2018
Citation: Covarrubias-Pazaran, G., Schlautman, B., Diaz-Garcia, L., Grygleski, E., Polashock, J.J., Johnson-Cicalese, J., Vorsa, N., Iorizzo, M., Zalapa, J.E. 2018. Multivariate GBLUP improves accuracy of genomic selection for yield and fruit weight in biparental populations of Vaccinium macrocarpon Ait. Frontiers in Plant Science. 9:1310. https://doi.org/10.3389/fpls.2018.01310.
DOI: https://doi.org/10.3389/fpls.2018.01310

Interpretive Summary: Recently, genetic and breeding applications that use massive genomic data have become feasible in major and minor species due to advances in next generation sequencing technologies. This manuscript presents the use of genetic (molecular markers) and trait information in cranberry breeding populations to test several variables important for prediction ability (i.e., predicting the performance of individuals based on genomic/trait information). We demonstrate that by using multiple variable methods, we increased predictive ability and obtained better resolution to detect genetic factors responsible for traits. In addition, we found that the use of optimal molecular marker densities and close genetic relationships between populations used played an important role in the magnitude of the predictive ability. This information will benefit future breeding efforts in American cranberry (Vaccinium macrocarpon) and other fruit crops.

Technical Abstract: The development of high-throughput genotyping has made genome-wide association (GWAS) and genomic selection (GS) studies possible for both major and minor crops. The exploitation of genomic techniques could greatly benefit future breeding efforts in American cranberry (Vaccinium macrocarpon) and other minor crops. Therefore, using cranberry breeding populations with different degrees of relationships, we compared univariate and multivariate quantitative trait loci (QTL) mapping and GS methods for total yield (TY) and mean fruit weight (MFW). In addition, we compared the differential predictive ability (PA) between traditional marker linear regression (MLR) against univariate and multivariate genomic best linear unbiased predictor (GBLUP and MGBLUP). We found GWAS methods to be more accurate to estimate the percent of variation explained by genetic components compared to single and composite QTL mapping. We found GBLUP to provide higher PA than MLR, and MGBLUP was not statistically different than GBLUP in all scenarios. Linkage disequilibrium decayed at approximately 18 cM (at r2=0.2), and only few hundreds of single nucleotide polymorphism (SNP) markers were needed to reach a plateau in PA. In addition, we found that higher resemblance among individuals in the training (TP) and validation (VP) populations provided greater PA. As better phenotyping and crop modeling techniques are developed, GS methods will quickly integrate into cranberry and other fruit breeding programs.