Bin-based genome-wide association studies reveal superior alleles for improvement of appearance quality using a 4-way MAGIC population in rice

Abstract

Introduction: The multiparental population provides us the chance to identify superior alleles controlling a trait for genetic improvement. Genome wide association studies at bin level (bin-GWAS) are expected to be more power in QTL mapping than GWAS at SNP level (SNP-GWAS).

Objectives: This study is to estimate genetic effects of QTL conferring grain appearance quality in rice by SNP-GWAS and bin-GWAS, compare their power in QTL mapping and identify the superior alleles of all detected QTL from 4 parents for genetic improvement.

Methods: A 4-way MAGIC population and its four founders were cultivated in two environments to dissect the genetic basis of rice grain appearance quality. Both SNP-GWAS and bin-GWAS were conducted for QTL mapping. Multiple comparison among 4 parental bin/alleles was used to identify the superior alleles.

Results: A total of 16 and 20 QTL associated with grain appearance quality were identified by SNP- and bin-GWAS, respectively. A minor chalkiness QTL qPGWC8.2/qDEC8 was assigned to a 30-kb genomic region, in which OsMH_08T0121900 is the potential candidate gene because its encoded protein, glucan endo-1,3-beta-glucosidase precursor is involved in the starch and sucrose metabolism pathway. The superior parental alleles for GS3, GL3.1, GW5, GW7, and Chalk5 and two QTLs were almost carried by the high-quality parents Cypress and Yuejingsimiao (YJSM), while the poor-quality parent Guichao-2 (GC2) always carried the inferior alleles. The top five recombinant inbred lines with the highest quality of grain shape and chalkiness traits all carried gene combinations of superior alleles.

Conclusions: Both SNP- and bin-GWAS methods are encouraged for joint QTL mapping with MAGIC population. qPGWC8.2/qDEC8 is a novel candidate gene strongly associated with chalkiness. The superior alleles of GS3, GW5, GL3.1, GW7, Chalk5 and qPGWC8.2 were identified, and the pyramiding of these superior alleles is helpful to improve rice appearance quality.

Keywords: Bin-GWAS; Chalkiness; Grain shape; Multiple comparisons; Superior allele.

Graphical abstract

Fig. 1

Distribution of different appearance quality traits in two years in the 4-way MAGIC population.

Fig. 2

Correlation heatmap between grain quality traits in the 4-way MAGIC population in 2015 and 2016. DEC: degree of endosperm chalkiness, PGWC: percentage of grains with chalkiness, GL: grain length, GW: grain width, and LWR: length–width ratio. The numbers 15 and 16 indicate the data from 2015 and 2016, respectively. The negatively correlated cells represented in blue, while the positively correlated cells represented in red. The darker color, the higher correlation was (P < 0.05; P < 0.01).

Fig. 3

Manhattan (A, C, E, G) and Q-Q (B, D, F, H) plots showing associated SNP markers for the degree of chalkiness (DEC) and percentage of grains with chalkiness (PGWC) detected by a GWAS at SNP level in 2015 and 2016. The X-axis shows the chromosome number, and the Y-axis shows −log10 (p), while the horizontal line indicates the threshold p-value at significant level (p < 0.0001). Chalk5 is indicated in the plots.

Fig. 4

Manhattan (A, C, E, G, I, K) and Q-Q (B, D, F, H, J, L) plots showing associated SNP markers for grain length (GL), width (GW) and length–width ratio (LWR) detected by a GWAS at the SNP level in 2015 and 2016. The X-axis shows chromosome number and the Y-axis shows −log10 (p), while the horizontal line indicates the threshold p-value at significance level (p < 0.0001). GS3, GL3.1, GW5 and GW7 indicated in the plots.