Linkage map construction
Genomic DNA was extracted from a single plant for each DH line and their parents (Zhang et al., 2008). SNP linkage maps were constructed for six DH populations, including two linkage maps of 90k SNP and four linkage maps of 12k Targeted Genotyping-By-Sequencing (tGBS). SNP genotyping was performed using an InfiniumTM iSelect assay on an Illumina iScan instrument according to the manufacturer’s protocols (Illumina, San Diego, CA). SNP clustering and genotype calling was performed using GenomeStudio v2011.1 software (Illumina, San Diego, CA) with the custom genotype-calling algorithm described by Cavanagh group (Cavanagh et al., 2013). Identical lines were detected and removed using non-metric multidimensional scaling (MDS) of genetic dissimilarity using software from Numerical Taxonomy System (NTsys) v2.2 and Plymouth Routines in Multivariate Ecological Research (PRIMER v6) (Clarke & Gorley, 2006; Rohlf, 2009).
Gene-based markers were used in conjunction with SNPs to account for the fact that in spring wheat, anthesis is highly controlled by the vernalisation genes (Zhang et al., 2014) and seed number per spike is strongly associated with the Rht1 gene (Zhang et al., 2013). Markers for Rht-B1b , Rht-D1b, VrnA1a ,VrnB1a , and VrnD1a were used to construct the final maps of the six DH populations using Map Manager (Manly et al., 2001) and the QTL mapping package R/qtl (Broman & Sen, 2009). Primers for VRN1(Vrn-A1a , Vrn-B1a and Vrn-D1a ), and Rht-B1 ,Rht-D1 were as described (Zhang et al., 2013; Zhang et al., 2014).
During the mapping process, a large proportion of identical lines were detected and removed. The high number of identical DH lines is probably due to the method of anther (microspore) culture used for DH line generation.
Lines with large proportion of missing values on SNP genotyping were also removed, together with distorted markers and many double-cross markers. Most co-segregating markers were made redundant and removed from the genetic map. In the St3B population, there were 163 identical lines among a total of 350 DH lines. A fine map of the St3B population was constructed using 185 DH lines and 1924 SNP markers. In the G7A population, 31 identical lines, 42 non-parental lines and 18 F1 lines were detected among a total of 327 DH lines. Excluding 34 samples with many missing data, 218 DH lines and 3592 SNP markers were used to construct a fine map of the G7A population. In the BW population, a total of 15 identical lines were detected among 105 DH lines. After excluding 12 DH lines with low calling rate, 77 lines and 2387 SNP markers were used to fine-map the BW population. In the Sp7A population, 48 identical lines were identified among 304 DH lines. After excluding 65 lines with low calling rate, 191 DH lines and 2367 SNP markers were used to generate the Sp7A map. A total of 12 identical lines were found among 168 DH lines in the SpB population. After excluding 62 lines with low calling rate, 94 DH lines and 2570 SNP markers were used to construct the linkage map for the SpB population. A relatively small proportion of identical lines (7) were detected among the 222 DH lines of the SpM population. After excluding 27 DH lines with low calling rate and high Heterozygous rate, 188 DH lines and 2235 SNP markers were used to construct the linkage map for SpM population (Supplementary Table 2).