Genetic dissection of grain elements predicted by hyperspectral imaging associated with yield-related traits in a wild barley NAM population.

Abstract

Enhancing the accumulation of essential mineral elements in cereal grains is of prime importance for combating human malnutrition. Biofortification by breeding holds great potential for improving nutrient accumulation in grains. However, conventional breeding approaches require element analysis of many grain samples, which causes high costs. Here we applied hyperspectral imaging to estimate the concentration of 15 grain elements (C, B, Ca, Cd, Cu, Fe, K, Mg, Mn, Mo, N, Na, P, S, Zn) in high-throughput in the wild barley nested association mapping (NAM) population HEB-25, comprising 1,420 BC 1 S 3 lines derived from crossing 25 wild barley accessions with the cultivar 'Barke'. Nutrient concentrations varied largely with a multitude of lines having higher micronutrient concentration than 'Barke'. In a genome-wide association study (GWAS), we located 75 quantitative trait locus (QTL) hotspots, whereof many could be explained by major genes such as NO APICAL MERISTEM-1 (NAM-1) and PHOTOPERIOD 1 (Ppd-H1). The GWAS approach revealed exotic alleles that were able to increase grain element concentrations. Remarkably, a QTL linked to GIBBERELLIN 20 OXIDASE 2 (HvGA20ox 2 ) significantly increased several grain elements without yield loss. We conclude that introgressing promising exotic alleles into elite breeding material can assist in improving the nutritional value of barley grains.

Plant phenotyping relevance

ハイパースペクトル画像から穀粒中15元素濃度を推定する高スループット表現型取得法を大規模集団に適用しており、画像による形質推定が研究の主要な技術的基盤です。

Code and data availability

The supplied blocks describe the paper's hyperspectral imaging phenotyping pipeline (5607 grain scans, Matlab workflow, PLS/RBF models) and GWAS, but contain no public dataset deposit, no author code/model availability statement, and no public URL for any paper-specific asset. The only URL present is the CC BY-NC-ND 4.

No evidence-backed public reproduction asset is currently recorded.