Fig. 1

GNP1 haplotype analysis. a The gene structure of GNP1. The coding region is indicated as black box, 3′ UTR and 5′ UTR are indicated as white boxes, and single nucleotide polymorphism (SNP) positions are connected to the haplotype table by lines (SNP frequency > 5%). Six potential intervals of promoters, marked with red stick, are found in the promoter region by promoter prediction software. Fourteen haplotypes, classified by SNP and insertion/deletion (InDel) variations, are detected in the collection (indicated as either blue or white). Numbers show the length of the InDel in bp. The number and subpopulation identity of varieties carrying each haplotype are indicated in columns to the right. The ten wild rice (O.rufipogon) varieties are indicated by WR1–9. b The phylogenetic network of GNP1. The phylogenetic network of haplotypes in Oryza sativa and O. rufipogon are constructed for GNP1. Each circle represents an allele. The size of the circles indicates the frequency of each allele. The green, yellow and blue circles indicate the allelic distribution in the wild rice, xian and geng species, respectively. The red median vectors (mv) represent a hypothesized sequence that is required to connect the existing sequences within the network with maximum parsimony. The solid lines represent one mutation step that interconnects adjacent alleles. c–d Grain number per panicle (c), plant height (d) are compared between the GNP1–3 and GNP1–4 haplotypes in geng rice (Oryza sativa). e–f Grain number per panicle (e), plant height (f) are compared among the GNP1–1,GNP1–2,GNP1–5 and GNP1–6 haplotypes in xian rice (Oryza sativa). Letters on histograms (a, and b) are ranked by Duncan’s test at P < 0.05. Pink, light yellow and light blue colors indicate in 2014, 2015 and 2016, respectively