Whole-genome sequencing (WGS) and assembly

To generate a D. rotundata genome sequence, an individual plant, TDr96_F1, was selected from the progeny in the open-pollinated D. rotundata breeding line TDr96/00629 (Fig. 1a, b). As TDr96_F1 never flowered during the current study period, we could not determine its sex. While D. rotundata is characterized by different ploidy levels (2× and 3×) with a basic chromosome number of 20 [8, 9], we found TDr96_F1 to be diploid (2n = 2× = 40) based on the mitotic chromosome number within root meristem cells (Fig. 1c). We estimated the genome size of TDr96_F1 to be 570 Mb by flow cytometry (FCM) analysis (Fig. 1d).

We assessed the quality of our assembly by investigating the presence of 248 highly conserved core eukaryotic genes with the Core Eukaryotic Genes Mapping Approach (CEGMA) [14] and confirmed the presence of 243 (98%) of those genes (Additional file 1: Table S4). Similarly, 94% of 956 Benchmarking Universal Single-Copy Orthologs (BUSCOs) [15] were present in at least one complete single copy in the assembly (Additional file 1: Table S5). Since the TDr96_F1 reference genome was generated from total genomic DNA, it also contained organelle-derived sequences. Alignment of the TDr96_F1 PE reads to the published D. rotundata chloroplast genome sequence [16] showed that 14.7% of the total PE reads were derived from the chloroplast genome (Additional file 1: Table S6). We also isolated mitochondrial DNA from TDr96_F1 leaves, sequenced this DNA using PE reads with Illumina MiSeq, and generated a 564-kb de novo assembly comprising 76 scaffolds (Additional file 2: Figure S5). Among PE reads, 1.25% represented mitochondrial sequences.

Generation of pseudo-chromosomes by anchoring scaffolds onto a linkage map

We developed a genetic map of D. rotundata using 150 F1 individuals obtained from a cross between two heterozygous breeding lines, TDr97/00917 (P1, female) and TDr99/02627 (P2, male), using restriction site associated DNA (RAD)-tags as DNA markers [17] (Additional file 2: Figures S6, S7) and the pseudo-testcross method [18, 19]. We aligned RAD-tags to TDr96_F1 scaffold sequences and selected DNA markers heterozygous in P1 and homozygous in P2, as well as markers heterozygous in P2 and homozygous in P1, resulting in 1326 and 1272 markers for P1 and P2 heterozygous sites, respectively (Additional file 1: Table S7 and Additional file 2: Figure S8). We then calculated the recombination fraction (rf) between the RAD markers to generate linkage maps. If the pairwise rf value of two RAD markers on the same scaffold exceeded 0.25, the scaffold was divided halfway between the markers because they were likely misassembled (see explanation in Additional file 2: Figure S9 and Additional file 1: Table S8). Two linkage maps, P1-Map and P2-Map, were generated based on the segregation pattern of the selected markers in the F1 progeny (Additional file 2: Figure S10), to which D. rotundata scaffolds were anchored using the 100-bp DNA sequences of RAD-tags. We combined the two maps using shared scaffolds (Additional file 2: Figures S11, S12), which allowed the ~ 454-Mb sequence (representing 76.4% of the assembly) to be anchored onto 21 linkage groups (LGs) to construct chromosome-scale pseudo-molecules (Fig. 2 and Additional file 1: Table S9). Smaller LGs could not be unequivocally mapped; hence, 21 LGs were obtained, whereas 20 LGs are expected based on the basic chromosome number. We validated the quality of our assembly by comparing the pseudo-molecule sequence with newly sequenced PE sequence reads having an average insert size of ~ 100 kb obtained from the BACs. Among the 315 BAC clones for which sequences of both ends could be mapped onto the assembly, 265 (84.1%) had both pairs in the same scaffold in the correct orientation, with an average distance of 116 kb (Additional file 1: Table S10 and Additional file 2: Figure S13), confirming the quality of our assembly. We compared the de novo assembled scaffolds to linkage information about the RAD markers, finding that 75% of RAD markers on the same scaffolds had rf < 0.25, and 73.5% of the scaffolds were retained without the need for splitting (Additional file 1: Table S11). The remaining 26.5% of the scaffolds had an rf > 0.25 and were divided into two or more scaffolds to solve the inconsistency between assembly and linkage information.

Guinea yam gene prediction and comparative genomics

We predicted genes and transposons using the TDr96_F1 reference genome sequence. To construct reliable gene models, we followed the MAKER pipeline using RNA-seq data from 18 samples representing various D. rotundata tissues (Additional file 1: Tables S12, S13) and combined the data with publicly available expressed sequence tags (ESTs) and homologous protein sequences from related angiosperm species (Additional file 2: Figure S14). This resulted in the prediction of 26,198 genes (Table 1 and Additional file 3), 22,477 (85.8%) of which are supported by RNA-seq data.

We compared Guinea yam genome sequence metrics with those of Arabidopsis thaliana (dicot), Brachypodium distachyon (monocot), and Oryza sativa (monocot) (Table 1). Interestingly, the GC contents of the total genome and exons of protein-coding genes in Guinea yam were 35.8% and 44.1%, respectively; these values are close to those of Arabidopsis and much lower than those of the Poales species Brachypodium and Oryza (Table 1). We annotated an average of 6.03 exons and 4.03 introns per gene. Roughly half of the genome was represented by an interspersed sequence (274.5 Mb), a major component of which was long terminal repeat (LTR) sequences (135.7 Mb) (Table 1).

We identified 5557 D. rotundata genes with a 1:1:1:1 orthologous relationship to the high-quality B. distachyon, O. sativa, and A. thaliana gene models (Fig. 3a and Additional files 4, 5). This number was reduced to 2795 genes when we included Arecales (Elaeis guineensis, Phoenix dactylifera) and Zingiberales (Musa acuminata) in our analysis (Additional files 6, 7). We constructed a phylogenetic tree based on the alignment of 2381 orthologous protein-coding genes in the five monocotyledonous species (Fig. 3b). D. rotundata did not group with any species in the tree, including Musa of Zingiberales, Phoenix and Elaeis of Arecales, and Oryza and Brachypodium of Poales, suggesting that Dioscorea diversified independently from these taxa in monocotyledons.

For 12,625 D. rotundata genes, no orthologs or paralogs were found in B. distachyon, O. sativa, or A. thaliana, and 11,348 D. rotundata genes had no clear homologs in any of the six species shown in Fig. 3a and Additional file 8. Of these 11,348 genes without homologs, 3422 were expressed in tuber tissues, a tissue type not shared with the other species examined.

Non-redundant Gene Ontology (GO) terms “intracellular organelle”, “protein binding”, and “ion binding” were significantly enriched among D. rotundata genes that showed no orthology to the other species, but not among the conserved genes (Additional files 9, 10). D. rotundata genes without orthologs in the other species included 68 genes encoding proteins with lectin domains that are involved in defense against microbial pathogens, nematodes, and insects, accounting for 31% of the 216 lectin-coding genes functionally annotated in D. rotundata. Among the 12 subfamilies of lectins [20], the bulb-type lectin (snowdrop lectin; B-lectin) family contributed the largest share (110) of genes in D. rotundata (Additional file 1: Table S14). Phylogenetic analysis of the B-lectin genes in D. rotundata (110 genes; 51 unique), B. distachyon, O. sativa, and A. thaliana revealed two expansions of B-lectin genes in Dioscorea (Fig. 3c). The first expansion (blue band) consisted of 22 receptor-like serine/threonine-protein kinases, which are thought to play a role in signaling and the activation of plant defense mechanisms [21]. The second expansion (red band) consisted of 28 mannose-binding lectins sharing high similarity with Dioscorea batatas tuber lectin DB1 (accession number AB178475). DB1 has insecticidal properties against cotton bollworm (Helicoverpa armigera), and studies in transgenic tobacco and rice plants expressing DB1 demonstrated that it also confers resistance against green peach aphid and brown plant hopper, respectively [22–24]. Of these mannose-binding lectin genes in Guinea yam, 16 did not have orthologs in any of the six other species examined, and two showed enriched expression (Benjamini–Hochberg [25] adjusted P value [padj] < 0.05) in tubers.

RNA-seq analysis comparing three tuber tissues to all other nine tissues (Additional file 1: Table S12) revealed that 2023 genes were enriched in tubers. The top 50 highly expressed (padj < 0.05) genes included genes encoding starch synthases and branching enzymes, as well as three carbonic anhydrase-encoding genes. Basic Local Alignment Search Tool (BLASTP) (https://blast.ncbi.nlm.nih.gov) analysis showed that these carbonic anhydrase-encoding genes shared high identity (average 76%) with genes encoding Dioscorea japonica precursors of dioscorin, a tuber storage protein that has carbonic anhydrase activity and exists in multiple isoforms [26] (Additional file 11).

To infer the past genome duplication in D. rotundata, we performed genome-wide dot plot analysis of D. rotundata against itself (Additional file 2: Figure S15), which revealed no indication of genome duplication. Nevertheless, we observed 946 paralogous gene clusters composed of duplicated genes in D. rotundata. Of these, 145 duplicate clusters of paralogous genes were observed only in D. rotundata. To investigate macrosynteny between D. rotundata and related species, we carried out whole-genome syntenic dot plot analysis against the genomes of Oryza sativa, Spirodela polyrhiza, and Phoenix dactylifera. At the chromosomal level, it was difficult to observe synteny conservation between these species. To assess microsynteny conservation, we performed a syntenic path assembly [27] of the scaffolds from these species against D. rotundata-masked pseudo-chromosomes (see Methods). The reordering and reorientation of the scaffolds relative to D. rotundata pseudo-molecules identified large proportions of the genomes to be conserved at the microsyntenic level (Additional file 2: Figure S16). This suggested that the D. rotundata genome has undergone many recombination events after its divergence from the other species.

Whole-genome resequencing of F1 bulk segregants identifies a genomic region associated with sex determination in D. rotundata

We previously developed a next generation sequencing (NGS)-based method for bulked segregant analysis (BSA) for quantitative trait locus (QTL) mapping in rice, named QTL-seq [28]. To our knowledge, this method has not been applied in species with highly heterozygous genomes. The majority of Dioscorea species, including D. rotundata, are mostly dioecious, with separate male and female plants (Fig. 4a), making it interesting to understand the genetic mechanism of sex determination in this genus. From a cross between two D. rotundata breeding accessions, TDr97/00917 (P3, female) and TDr97/00777 (P4, male), we generated an F1 population of 253 individuals in 2014 that segregated for male, female, monoecious (male and female flowers on the same plant), and non-flowering types (Additional file 1: Table S15). For QTL-seq analysis (see Additional file 2: Figure S17 for details), we sequenced two DNA bulks representing male and female plants, each from 50 individuals, generating 7.9- and 7.3-Gb sequences, which provided 13.9× and 12.7× coverage of the predicted D. rotundata genome, respectively (Additional file 1: Table S16). We also resequenced the genome of the female parent (P3) and generated a P3 reference sequence (P3-Ref) by replacing TDr96_F1 nucleotides with P3 nucleotides at all different sites between the two genotypes. Likewise, we generated the male parent (P4) reference sequence (P4-Ref) by aligning P4 sequence reads to TDr96_F1 and replacing TDr96_F1 nucleotides with those of P4 at all different sites. We then separately aligned sequence reads obtained from F1 male-bulk and female-bulk DNA to the P3- and P4-Ref sequences. To identify single-nucleotide polymorphism (SNP) markers associated with the F1 gender phenotype, thus potentially suggesting candidate sex-determining gene(s), we focused on SNPs that segregated in the F1 progeny either as SNPs homozygous in the female parent (P3) but heterozygous in the male parent (P4), or vice versa. We could then identify genomic regions with SNPs heterozygous in one parent whose alleles were differentially transmitted to the two sexes in the F1, suggesting Y or W linkage, respectively. This is similar to mapping by backcrossing, but does not require a BC1 generation using inbred lines. Scanning the entire genome identified a single region, from 0.65 Mb to 2.35 Mb on pseudo-chromosome 11, whose SNP-index values (the frequency of short reads aligned to a particular position of the genome with SNPs different from the reference sequence [28]) differed for male and female bulks in the second category of SNPs just described (Fig. 4b and Additional file 2: Figure S18).

We identified a sex-linked region with category 2 SNP markers that are heterozygous in the female parent (P3) but homozygous in the male parent (P4) (Additional file 2: Figure S18), suggesting that the male sex is determined by the homozygous (designated ZZ) state of the locus responsible for sex determination, whereas that of the female sex is determined by the heterozygous (ZW) (or hemizygous: Z-) state of this locus (Fig. 4c). Genotyping of the F1 individuals used for bulk sequencing using the cleaved amplified polymorphic sequence (CAPS) marker sp1 developed within the candidate genomic region revealed significant co-segregation between the sp1 marker and the sex of the individual (P = 1.913e-14, Fisher’s exact test). This analysis confirmed that the genomic region identified by QTL-seq is indeed associated with sex determination (Fig. 5a, b and Additional file 2: Figure S19). The switch of sp1 male and female marker genotypes in the F1 progeny occurred because the marker genotype was heterozygous in the female parent (Fig. 4c).

As the TDr96_F1 plant never flowered, we were unable to determine its sex based on flower phenotype and therefore could not directly characterize its genotype (ZZ or ZW) at the candidate sex locus. To identify the genomic regions linked to Z and W, we assembled the P3 (female) and P4 (male) genomes de novo using their PE reads with the DISCOVAR De Novo assembler [29], generating P3-DDN (female, N50 = 3.3 kb) and P4-DDN (male, N50 = 2.7 kb) reference sequences (Fig. 6a, Additional file 1: Table S17 and Additional file 2: Figure S20). We separately mapped short reads derived from bulked DNA from 50 male and 50 female F1 progeny (P3 × P4) to P3-DDN and P4-DDN and looked for unique P3-DDN (female) genomic regions (presumably corresponding to the W-linked region) that were specifically mapped by F1 female-bulk reads but not by F1 male-bulk reads. The 1345 regions (sizes from 1 to 129 bp) totaling 15,390 bp conformed to this pattern (Additional file 2: Figure S21). We ordered these fragments by size and found that the N20 value was 42 bp. Conversely, we found only 435 regions (total size 3775 bp) of P4-DDN (male) mapped by the F1-male bulk but not by the F1-female bulk (Additional file 2: Figure S21). The large size difference between female-specific P3-DDN regions (total 15,390 bp) and male-specific P4-DDN regions (total 3775 bp) suggested that the ZW female genome has additional DNA sequences not present in the ZZ male. We hypothesize that the recovery of small male-specific P4-DDN regions may have occurred by chance. We focused on 36 female-specific contigs of P3-DDN that contained DNA fragments larger than 42 bp (Fig. 6b and Additional file 2: Figure S21). When we used the 36 contigs as BLASTN queries against the TDr96_F1 reference genome, 20 were located on scaffold206 (667.8 kb) on pseudo-chromosome 11 (Fig. 6c, Additional file 1: Table S18), suggesting that P3-DDN contigs with female-specific regions were indeed located within the sex-linked region identified by QTL-seq (Fig. 4b). We developed a PCR primer pair for one such P3-DDN contig (Fig. 6b; Female917_flattened_line_87512_3057) harboring female-specific regions; we named this DNA marker sp16. sp16 amplified a PCR fragment in the P3 female parent but not in the P4 male parent (Fig. 7a), demonstrating that this fragment was located in the female-specific region. An sp16 PCR fragment was amplified in TDr96_F1, our reference genome plant (Fig. 7a), suggesting that this individual likely had the ZW genotype. In F1 progeny derived from a P3 × P4 cross, the sp16 fragment was amplified in all female plants, but it failed to be amplified in the majority of male individuals. Furthermore, sp16 fragments were amplified in monoecious as well as non-flowering progeny (Fig. 7a). We monitored flowering in all 249 F1 individuals in two consecutive seasons (2014 and 2015) and found that 194 plants showed consistent sex phenotypes. However, the remaining 55 plants showed changes in sex among male, female, and monoecious (Fig. 7b). Genotyping of all F1 individuals using sp16 revealed a striking pattern: 121 of the 125 plants that were consistent for male over the 2 years showed no PCR amplification of sp16, whereas all plants with the remaining phenotypes showed amplification of sp16 (Fig. 7b). A similar pattern was observed in another F1 family (TDr04-219 × P4) involving the same male parent, P4 (Fig. 7c). We also assayed 24 Guinea yam breeding accessions of known sex using the same marker (Fig. 8). All 10 female accessions, as well as three accessions that did not flower, showed amplification of sp16. Of the 11 male accessions genotyped, eight did not show amplification of sp16, whereas the remaining three did.

We hypothesized that the ZZ genotype stably gives rise to the male phenotype, whereas the ZW genotype results in unstable sex phenotypes; ZW mainly generates the female phenotype, but sometimes monoecious or male phenotypes depending on the environments. Therefore, some individuals of the F1 progeny derived from a cross between P3 and P4 might have been scored as male despite their genotype being ZW, which may have obscured our analysis, resulting in non-zero depth of male DNA bulk within the putative W-region (Fig. 6b). To address this possibility, we selected 50 ZZ plants from the F1 progeny based on their sp16 genotype and bulked and sequenced the DNA (sp16-minus bulk). The sp16-minus bulk reads, as well as female bulk reads, were separately mapped to the combined sequence of the TDr96_F1 reference genome and P4-DDN to identify the female-specific TDr96_F1 genomic region, as described in Fig. 6a. As shown in Fig. 6c and Additional file 2: Figure S20c, d, we successfully delineated the putative W-linked region mapped predominantly with female-only bulk DNA, representing an approximately 161-kb region of scaffold206 on pseudo-chromosome 11. This putative female-specific W-linked region contains ~ 10 predicted genes (Additional file 12).

Plant materials

The TDr96_F1 line used for WGS was selected from F1 progeny obtained from an open-pollinated D. rotundata breeding line (TDr96/00629) grown under field conditions in the experimental fields of the International Institute of Tropical Agriculture (IITA) in Nigeria. F1 seeds from TDr96/00629 and those obtained from the cross between the parental lines TDr97/00917 and TDr99/02627 used for RAD-seq were germinated on wet paper towels in darkness at 28 °C. After germination, the seeds were transferred to soil (Sakata Supermix A [34]) and grown at 30 °C with a 16-h/8-h photoperiod in a greenhouse at Iwate Biotechnology Research Institute (IBRC) in Japan. Fresh leaf samples were collected for DNA extraction. Additionally, to resequence the F1 progeny used for QTL-seq analysis, lyophilized leaf samples obtained from plants that were grown and phenotyped under field conditions at IITA were used for DNA extraction.

Determination of chromosome number and ploidy level

For chromosome observation, root tips of TDr96_F1 plants generated by in vitro propagation of nodal explants were sampled and fixed in acetic acid-alcohol (1:3 ratio) for 24 h without pretreatment. Fixed root tips were stained with a 1% aceto-carmine solution for 24 h. Samples were prepared by the squash method and analyzed under an Olympus BX50 optical microscope (Olympus Optical Co, Ltd., Tokyo, Japan [35]) at 400× magnification.

Estimation of D. rotundata genome size

Whole-genome sequencing

For WGS, genomic DNA was extracted from fresh TDr96_F1 leaf samples using a NucleoSpin Plant II Kit according to the manufacturer’s protocol (Macherey-Nagel GmbH & Co. KG [38]) with slight modifications. Homogenized samples were washed with 0.1 M 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffer to remove contaminating polysaccharides. Just before use, 120 mg polyvinylpyrrolidone (PVP), 90 mg l-ascorbic acid, and 200 μl 2-mercaptoethanol (ME) were added to 10 ml HEPES buffer, and 1 ml of the mixture was used to wash each sample; washing was repeated three times. Additionally, 10 μl 2-ME and 5 μl of 30% polyethylene glycol (PEG)-20000 were added to 1 ml of PL1 buffer (provided with the NucleoSpin Plant II Kit), and twice the recommended volume of buffer (800 μl) was used for cell lysis. Libraries for PE short reads and MP jump reads of various insert sizes including 2, 3, 4, 5, 6, and 8 kb were constructed using an Illumina TruSeq DNA LT Sample Prep Kit and a Nextera Mate Pair Sample Prep Kit, respectively. The PE library was sequenced on the Illumina MiSeq platform, while the MP libraries were sequenced on the HiSeq 2500 platform. Library construction and sequencing of the 20- and 40-kb MP jump sequences were carried out by Eurofins Genomics (Operon [39]) and Lucigen [40], respectively. The 20-kb and 40-kb jump libraries were sequenced on the MiSeq and HiSeq 2500 platforms, respectively. BAC libraries were constructed by Lucigen, and BAC-end sequencing was carried out by Genaris [41] using Sanger sequencing. A total of 30,750 clones corresponding to 3072 Mb of sequence and 5.4× genome coverage were constructed. Of these, 9984 clones were used for BAC-end sequencing, generating a 13.6-Mb sequence in PE fasta format, which was converted to 50-bp PE short reads corresponding to a 0.46-Gb sequence and ~ 0.8× coverage of the estimated 470-Mb D. rotundata genome (Additional file 1: Table S2 and Additional file 2: Figure S2).

De novo assembly

All TDr96_F1 sequence reads in fastq format were filtered for quality using the FASTX-Toolkit version 0.0.13 [42]. For PE reads and MP short jump reads with insert sizes ranging from 2 to 8 kb, only those having sequence reads with a Phred quality score of ≥ 30 (i.e., ≥ 90% of the reads) were retained. Adapter trimming and removal of MP reads with the wrong insert sizes were performed using an in-house pipeline of scripts written in Perl and C++. Quality filtering of the long jump sequences (20-, 40-, and 100-kb insert sizes) was carried out by the suppliers. For the initial de novo assembly, short PE reads and MP jump reads with 2- to 40-kb insert sizes (Additional file 1: Table S2) were assembled using ALLPATHS-LG assembler version R49856 [11]. Further scaffolding of the assembly generated by ALLPATHS-LG was performed using the 100-kb jump MP fastq reads obtained by BAC-end sequencing and the SSPACE PREMIUM 2.3 scaffolding tool with default parameters [13].

Constructing organelle genome sequences

De novo assembly of the D. rotundata mitochondrial genome sequence was performed using mitochondrial DNA isolated from TDr96_F1 leaf samples according to the method of Terachi and Tsunewaki [43] with the following minor modifications. Fresh green leaves (ca. 150 g) were homogenized in 1.5 L of homogenization buffer containing 0.44 M mannitol, 50 mM Tris-HCl (pH 8.0), 3 mM ethylenediaminetetraacetic acid (EDTA), 5 mM 2-ME, 0.1% (w/v) bovine serum albumin, and 0.1% (w/v) PVP. Following DNaseI treatment, the mitochondrial fraction was collected from the interface between 1.30 M and 1.45 M of a sucrose gradient. Mitochondrial DNA was purified by EtBr/CsCl centrifugation at 80,000 rpm for 6 h at 20 °C in a Beckman TLA 100.3 rotor. The DNA band was collected and purified by ethanol precipitation. The resulting mitochondrial DNA (15 ng) was amplified using a REPLI-g Mini Kit (Qiagen, Cat. no. 150023) and used for library construction. The library was sequenced on an Illumina MiSeq sequencer, and the resulting PE reads were assembled de novo using DISCOVAR De Novo [29], generating D. rotundata mitochondria contigs. For scaffolding, MP reads with insert sizes of 2, 3, 4, 5, 6, 8, and 20 kb obtained from D. rotundata genomic DNA (gDNA) were aligned to the D. rotundata mitochondrial contigs. MP reads showing 100% alignment were selected and used for scaffolding of D. rotundata mitochondrial contigs by SSPACE [13] (Additional file 2: Figure S5). To reconstruct the D. rotundata chloroplast genome sequence, the PE reads of TDr96_F1 were aligned to the recently published D. rotundata chloroplast genome sequence [16] (GenBank ID = NC_024170.1) by Burrows-Wheeler alignment (BWA) [44], and chloroplast-derived sequences were identified, amounting to 5,403,420 reads (14.74% of the total size of PE reads generated for TDr96_F1 [Table 1]) matching the assembled 155.4-kb chloroplast genome of D. rotundata.

Evaluation of the completeness of the genomic assembly

To evaluate the completeness of the D. rotundata genome assembly, the assembly was checked for the presence of 248 highly conserved core eukaryotic genes [45] using CEGMA version 2.4 with default parameters [14] (Additional file 1: Table S4). To further assess the completeness of the genome, the successor to CEGMA, Benchmarking Universal Single-Copy Orthologs (BUSCO), was used to check for the presence of 956 BUSCOs with version 1.1.b1 [15] using the early access plant dataset (Additional file 1: Table S5).

Annotation of transposable elements (TEs)

Legacy repetitive sequences, including transposons, were predicted using CENSOR 4.2.29 [46] with the following options: show_simple, nofilter, and mode rough using the Munich Information Center for Protein Sequences (MIPS) Repeat Element Database [47]. Following identification, the repeat elements were classified using mips-REcat [47]. Repetitive sequences were later improved by remodeling using RepeatModeler 1.0.8 [48] and masked with RepeatMasker 4.0.5 [49]. Using the National Center for Biotechnology Information (NCBI) database, one of three other options was used to generate interspersed RepeatModeler-based, interspersed Rebase-based, and Low complexity repeats: “nolow”, “nolow, species Viridiplantae”, and “noint”, respectively. Repeat element content and other statistics were compared between the D. rotundata and A. thaliana TAIR10 [50], B. distachyon v3.1 [51], and O. sativa v7_JGI 323 [52] genomes using the RepeatModeled and RepeatMasked references (Table 1).

RNA-seq

Total RNA was extracted using leaf, stem, flower, and tuber samples collected from a greenhouse-grown TDr96_F1 plant using a Plant RNeasy Kit (Qiagen [53]) with slight modifications. RLC buffer was used for lysis after the addition of 5 μl 30% PEG-20000 and 10 μl 2-ME to 1 ml of buffer. The RNA samples were treated with DNase (Qiagen) to remove contaminating genomic DNA. Two micrograms of total RNA was used to construct complementary DNA (cDNA) libraries using a TruSeq RNA Sample Prep Kit V2 (Illumina) according to the manufacturer’s instructions. The libraries were used for PE sequencing using 2× 100 cycles on the HiSeq 2500 platform in high-output mode. Illumina sequencing reads were filtered by Phred quality score, and reads with a quality score of ≥ 30 (≥90% of reads) were retained (Additional file 1: Table S12). Only one RNA-seq experiment was carried out per tissue/organ (indicated as sample in Additional file 1: Table S12).

Prediction of protein-coding genes

The legacy gene models were generated previously using the legacy repeat-masked reference genome and three approaches: ab initio, ab initio supported by evidence-based prediction, and evidence-based prediction. The ab initio prediction was carried out with FGENESH 3.1.1 [54]. The ab initio supported by evidence-based prediction was performed with AUGUSTUS 3.0.3 [55] using the maize5 training set and a hint file as the gene model support information. To construct the hint file, TopHat 2.0.11 [56] was used to align RNA-seq reads from tuber, flower (young), leaf (young), stem, leaf (old), and flower (old) samples to the D. rotundata reference genome, and Cufflinks 2.2.1 [57] was used to generate gene models from these data. The evidence-based predictions using the Program to Assemble Spliced Alignments (PASA) [58] were generated in a Trinity [59] assembled transcriptome from the RNA-seq data. JIGSAW 3.2.9 [60] was used to select and combine the gene models obtained using the three approaches with the weighting values assigned to the results from FGENESH, AUGUSTUS, and PASA of 10, 3, and 3, respectively. In total, 21,882 consensus gene models were predicted. These gene models were further improved upon using the MAKER [61] pipeline (Additional file 2: Figure S14). Publicly available ESTs and protein sequences from related plant species were aligned to the genome using GMAP [62] and Exonerate 2.2.0 [63], respectively. De novo and reference-guided transcripts were assembled from RNA-seq data from all 18 tissues using Bowtie 1.1.1 [64], Trinity 2.0.6 and SAMtools 1.2.0 [65], and Trinity 2.0.6 and TopHat 2.1.0, respectively. Both sets of assembled transcripts were used to build a comprehensive transcript database using PASA (Additional file 1: Table S13). High-quality non-redundant transcripts from PASA were used to generate a training set for AUGUSTUS 3.1. Gene models were predicted twice using the genome, improved repeat sequences, assembled transcripts, EST and protein alignments, the AUGUSTUS training set, and a legacy set of 21,882 gene models obtained previously using MAKER 2.31.6 [61], retaining all legacy gene models or querying them with new evidence and discarding those that could not be validated. From both MAKER runs, 21,894 and 76,449 gene models were predicted, respectively. A consensus set of gene models from both MAKER outputs was obtained using JIGSAW 3.2.9 [60] at a 1:1 ratio. In total, 26,198 consensus gene models were predicted in the D. rotundata genome. The corresponding amino acid sequences were also predicted for these gene models. To confirm these gene models, the RNA-seq reads were aligned to the CDSs (coding sequences) of the predicted genes using BWA [44] with default parameters. Accordingly, 85.8% of the gene models could be aligned by at least a single RNA-seq read. Functional annotation of the amino acid sequences was performed using the in-house pipeline, AnnotF, which compares Blast2GO [66] and InterProScan [67] functional terms.

Comparative genomics

Pairwise orthology relationships were determined with Inparanoid [68–70] using the longest protein-coding isoform for each gene in Arabidopsis thaliana (TAIR10) [50], Oryza sativa japonica (v7.0) [52], Brachypodium distachyon (v3.1) [71], Musa acuminata (v2) [72], Elaeis guineensis (EG5) [73], and Phoenix dactylifera (DPV01) [74]. Orthology clusters across all seven species were determined using Multiparanoid [75]. Sequences for the 12 classes of lectins were obtained from UniProt [76] for the proteomes of A. thaliana (up000006548), B. distachyon (up000008810), and O. sativa (up000059680). Protein alignments for B-lectin class protein sequences from all three of these species and D. rotundata were generated using the program Multiple Alignment using Fast Fourier Transform (MAFFT) [77]. Maximum likelihood trees were constructed based on the concatenated alignments of all 378 B-lectin proteins using RAxML [78] 8.0.2 with 1000 bootstraps. Enrichment of tuber-specific genes was detected using TopHat 2.1.0 to align RNA-seq data from each of the 12 tissues to the genome, with one biological replicate for each tissue. HTSeq 0.6.1 [79] was used to generate raw counts. Then the Bioconductor package DESeq2 1.14.1 [80] was used to compare raw counts of the three tuber tissues against all the other nine tissues (Additional file 1: Table S12) to determine tuber-enriched gene expression based on a log2 fold change > 0 and Benjamini–Hochberg [25] adjusted P value < 0.05.

Gene enrichment analysis of orthology clusters was performed with GOATOOLS [81], using the Holm significance test, and the false discovery rate was adjusted using the Benjamini–Hochberg procedure [25]. The list of enriched genes was filtered for redundant Gene Ontology (GO) terms using REVIGO [82]. For the species phylogeny, protein alignments for each gene with a 1:1 orthologous relationship across all monocot species were generated with MAFFT using the longest protein isoform. Maximum likelihood trees were constructed based on the concatenated alignments of 2381 orthologous protein-coding genes using RAxML 8.2.8 [78] with a JTT + Γ model and 1000 bootstraps.

SynMAP [83] using BLASTZ [84] alignments, DAGchainer [85] (options -D 30 and -A 2), and no merging of syntenic blocks were used as part of the CoGe platform [86] to identify syntenic blocks between the hard-masked pseudo-chromosomes of D. rotundata and scaffolds/contigs of Oryza sativa japonica (A123v1.0), Spirodela polyrhiza (v0.01), and Phoenix dactylifera L. (v3). A syntenic path assembly was then carried out on each of the same three species in SynMap using synteny between the scaffolds/contigs against D. rotundata pseudo-molecules. The syntenic path assembly is a reference-guided assembly that uses the synteny between two species to order and orientate contigs. This approach highlights regions of conservation that were otherwise too shuffled to be clearly observed. Self-self synteny analysis of D. rotundata pseudo-chromosomes was carried out using SynMap Last alignments with default parameters and syntenic gene pair synonymous rate change calculated by CodeML [87].

RAD-based linkage mapping and scaffold anchoring

RAD-seq was performed as previously described [88] with a minor modification. Genomic DNA was digested with the restriction enzymes PacI and NlaIII to prepare libraries used to generate PE reads by Illumina HiSeq 2500 (Additional file 2: Figure S6). Approximately 822.7-Mb and 250.4-Mb sequence reads covering 22.9% and 5.3% of the estimated 504-Mb D. rotundata genome sequence, excluding gap regions, at average depths of 7.2× and 9.8× were generated for the parental lines and F1 individuals, respectively (Additional file 2: Figure S7).

QTL-seq analysis

DNA samples obtained from the two parental lines, TDr97/00917 (P3, female) and TDr97/00777 (P4, male), as well as samples pooled in equal amounts from 50 male (male-bulk) and 50 female (female-bulk) F1 individuals obtained from the cross between P3 and P4 were subjected to WGS. Libraries for sequencing were constructed from 1-μg DNA samples with a TruSeq DNA PCR-Free LT Sample Preparation Kit (Illumina) and were sequenced via 76 cycles on the Illumina NextSeq 500 platform. Short reads in which more than 20% of sequenced nucleotides exhibited a Phred quality score of < 20 were excluded from further analysis. To perform QTL-seq analysis of F1 progeny, two types of analyses are required. In the first analysis, the SNP index and ΔSNP index are calculated at P4-specific heterozygous positions. The second analysis is performed using P3-specific heterozygous positions. To identify P4-specific heterozygous positions, the P3 “reference sequence” was first developed by aligning P3 reads to the reference genome sequence of D. rotundata and replacing nucleotides of the D. rotundata reference genome sequence with nucleotides of P3 at all SNP positions showing an SNP index of 1 (Additional file 2: Figure S17c). SNP detection, calculation of SNP index, and replacement of SNPs were carried out via step 2 of QTL-seq pipeline version 1.4.4 [90]. Short reads obtained from both the male and female parents were then aligned to the “reference sequence” and heterozygous SNP positions between the two were extracted. A SNP was defined as heterozygous if the same position showed an SNP-index value ranging from 0.4 to 0.6 in one parent and a value of 0 in the second parent. Of the selected markers/positions, only those having enough depth in both parents (>15) were used for analysis of SNP-index values in the bulk-sequenced samples. P3-specific heterozygous positions were identified similarly using the P4 “reference sequence.”

After identifying P4- and P3-specific heterozygous positions, the Illumina reads from the two bulk-sequenced samples (male and female bulks) were aligned to the reference sequences using BWA [44] and subjected to Coval filtering [91] as previously described. When the P3 reference sequence was used for alignment, the SNP-index values were calculated only at all of the P4-specific heterozygous positions. By contrast, when the P4 reference sequence was used for alignment, the SNP-index values were calculated only at the P3-specific heterozygous positions. In both cases, positions with shallow depth (< 6) in either of the two samples were excluded from analysis. The ∆SNP index was calculated by subtracting the SNP-index values of the male bulk from those of the female bulk. To generate confidence intervals of the SNP-index value, an in silico test simulating the application of QTL-seq to DNA bulked from 50 randomly selected F1 individuals was performed as described previously [28] (Additional file 2: Figure S22). The simulation test was repeated 10,000 times depending on the alignment depth of short reads to generate confidence intervals. These intervals were plotted for all SNP positions analyzed. Finally, sliding window analysis was applied to SNP-index, ∆SNP-index, and confidence interval plots with a 1-Mb window size and a 50-kb increment to generate SNP-index graphs (Additional file 2: Figure S18).

Identification of putative W-region by de novo assembly of female and male parental genomes and mapping of bulked DNA from female and male F1 progeny

DNA samples obtained from the two parental lines, TDr97/00917 (P3, female) and TDr97/00777 (P4, male), were separately subjected to de novo assembly. Libraries for sequencing were prepared with a TruSeq DNA PCR-Free LT Sample Preparation Kit (Illumina) and were sequenced for 251 cycles on the Illumina MiSeq platform. Contigs were generated using the DISCOVAR De Novo assembler [29], resulting in P3-DDN and P4-DDN, respectively. Separately, whole-genome resequencing of bulked DNA was performed on bulked DNA samples obtained from 50 female F1 (Female-bulk.fastq) and 50 male F1 (Male-bulk.fastq) progeny, all derived from a cross between P3 and P4. Two reference sequences, P3-DDN and P4-DDN, were combined to generate P3-DDN/P4-DDN. Short reads from the female and male bulks were separately mapped to P3-DDN/P4-DDN using the alignment software BWA [44]. After mapping, the MAPQ scores of the aligned reads were obtained. Under our conditions, if a short read was mapped to a unique position of the reference sequence, the MAPQ score was 60, whereas if the read was mapped to multiple positions, MAPQ was < 60. Since two reference sequences (P3-DDN and P4-DDN) were fused to generate P3-DDN/P4-DDN, most genomic regions were represented twice. Therefore, most short reads mapped to two or more positions, leading to a MAPQ score < 60. The reads that mapped to the P3-DDN/P4-DDN with MAPQ = 60 were judged to be located in either P3- or P4-specific genomic regions. After finding these P3- or P4-specific genomic regions, the depth of short reads that covered the regions for Female-bulk.fastq and Male-bulk.fastq, respectively, was evaluated. If the depth of Female-bulk.fastq was high and the depth of Male-bulk.fastq was 0 or close to 0, such genomic regions were retained as putative W-regions (Fig. 6 and Additional file 2: Figure S20).

DNA markers linked to sex

The primer sequences used for amplification of sex-linked markers sp1 and sp16, as well as the control Actin gene fragment (Dr-Actin), were as follows:

PCR was performed using the following conditions: 30 cycles of 98 °C for 10 s, 55 °C for 30 s, and 72 °C for 1 min. For CAPS analysis of the sp1 marker, the amplified DNA was digested with EcoRI. All PCR products were electrophoresed on 1.5% agarose gels.

Identification of SSR markers

Approximately 4,932,582 bp simple sequence repeat (SSR) motif-containing sequences were predicted in the D. rotundata genome. Within this region, SSR sequences with enough flanking regions were identified and evaluated for use in primer design. Accordingly, 134,101 SSR-containing sequences, excluding those with single base repeats, were identified. The SSR information for these sequences was analyzed using GMATo [92] version 1.2 Build 20130106 with the following parameters: m (minimum motif length) = 2, x (maximum motif length) = 10, and r (minimum repeat number) = 10. The necessary information was obtained for 22,164 SSR-containing sequences in the assembled genome, 12,724 (57.4%) of which were anchored to the genetic map (Additional file 1: Table S19). Primer pairs were designed for 1000 of these sequences using Primer3 [93] software release 2.3.6 with the following parameters: product size = 100–500, primer length = 18–22 bp (optimum 20 bp), GC content = 40–60% (optimum 50%), and Tm = 57–63 (optimum = 60.0).