UC Davis

Barley (Hordeum vulgare L.) is the fourth largest cereal crop produced in the world, and the majority of barley in the United States is malted for use in brewing/distilling. Variability in starch- and protein-related physical and compositional grain quality traits due to genotypic and environmental effects, and their interactions, can have important implications for the malting and brewing industry. Uncertainties in predicting barley grain quality can cause operational losses for farmers and maltsters, when inferior batches of grain are sourced that need protocol optimization during malting and brewing. This was the first examination of in-depth structural and compositional grain/malt quality traits in barley, using a contrasting set of 12 genotypes, grown in eight locations in California, across four crop years. The main effect of environment and/or the genotype-environment interaction explained the largest variance in yield, total protein content, grain size, endosperm texture, and starch-related traits. In assessing the proteome of these barley samples representing genotype-environment combinations, this study found the highest number of unique proteins to date (3105 proteins of which 828 proteins remain uncharacterised), with growing location explaining the largest variance in the overall proteome. Sixteen proteins with either storage, DNA/RNA binding, or enzymatic functionalities were found to be significantly high/low in relative abundance (fold change ± 2.0; false discovery rate of 0.01%; Padj<0.05 using Benjamini-Hochberg method) in a certain crop year, location, or genotype compared to the overall mean. These results collectively show that environment played a larger role in influencing the majority of starch- and protein-related physical and compositional traits, compared to genotype. Predictive models using individual protein abundances as predictors had reasonable accuracies for total protein content, alcohol-soluble protein content, malt protein content, and malt fine extract. During scenarios of high environmental variability, assaying in-depth traits like alcohol-soluble protein content, assessing the proteome to explore proteins that are significantly high/low in abundance, and predicting grain/malt quality traits using a combination of proteomics and machine learning could be invaluable to ensure sustainable supply.