Genomic prediction integrating genotype-by-environment interaction and environmental covariates for grain yield in unbalanced historical tropical maize-breeding data

Description of the topic, objectives

1) Get the analyses of the unbalanced maize historical data, collected across approximately 10 years and 14 locations, from Embrapa Maize and Sorghum breeding program in Brazil, is available for this study.

2) Use Environmental covariates collected during the crop growth cycle, such as mean, minimum and maximum daily temperature, daily precipitation, daily global radiation and daily relative humidity, are available for most of the years and locations.

3) Use the  SNP markers, obtained via genotyping-by-sequencing (GBS), are available for most of the parents of approximately 800 single-cross hybrids.

Work expectations

Fullfill the 3 main objectives described above. Prepare one or two scientific articles to be published in referee international journals.

Required skills

Basic skills in R software. Basic knowledge on statistical models.


Work with Jose Crossa, Paulino Perez, and other members of the BSU team to learn how to use the BGLR package for genomic selection. To learn how to use other biometrical genetic software developed by BSU for the analyses of its own data. Interact with maize breeders at CIMMYT