학술논문

Genome‐wide association mapping and comparative genomics identifies genomic regions governing grain nutritional traits in finger millet (Eleusine coracana L. Gaertn.)
Document Type
article
Source
Plants, People, Planet, Vol 2, Iss 6, Pp 649-662 (2020)
Subject
Finger millet
genome‐wide association analysis
genotyping‐by‐sequencing
micronutrients
minerals
population structure
Environmental sciences
GE1-350
Botany
QK1-989
Language
English
ISSN
2572-2611
Abstract
Societal Impact Statement Micronutrient deficiency is a serious and underestimated global health concern. Identifying existing micronutritional richness in traditional crops, and breeding this potential into staple crops that are more frequently consumed, could offer a potential low‐cost, sustainable solution to micronutrient deficiency. Here, we provide the first insight into genetic control of grain micronutrient content in the staple food crop finger millet (Eleusine coracana). Quantifying the existing natural variation in nutritional traits, and identifying the regions of the genome associated with these traits, will underpin future breeding efforts to improve not only global food and nutrition security, but also human health. Summary Finger millet is an excellent cereal crop to alleviate micronutrient malnutrition in marginal communities because of its nutrient‐dense grains. Identification of the alleles governing these characteristics will help to develop improved germplasm. An assembly of 190 genotypes was evaluated for six minerals (iron, zinc, calcium, magnesium, potassium, and sodium) and protein content. A combination of genotyping‐by‐sequencing (GBS) and genome‐wide association study (GWAS) was applied to identify marker‐trait associations (MTAs). Candidate genes underlying significant associations were predicted through comparative mapping with other cereals. A wide range of variation was observed for all traits. GBS generated 169,365 single‐nucleotide polymorphisms and three subpopulations were identified. GWAS, using general linear and mixed model approaches to correct for population structure and genetic relatedness, identified 418 common markers (p‐value ≤ 10–3, FDR