ABSTRACT A high-density linkage map is crucial for the identification of quantitative trait loci (QTLs), positional cloning, and physical map assembly. Here, we report the development of a

high-density linkage map based on specific length amplified fragment sequencing (SLAF-seq) for adzuki bean and the identification of flowering time-related QTLs. Through SLAF library

construction and Illumina sequencing of a recombinant inbred line (RIL) population, a total of 4425 SLAF markers were developed and assigned to 11 linkage groups (LGs). After binning the

SLAF markers that represented the same genotype, the final linkage map of 1628.15 cM contained 2032 markers, with an average marker density of 0.80 cM. Comparative analysis showed high

genomic research, such as identifying QTLs for other important traits, positional cloning, and comparative mapping in legumes. SIMILAR CONTENT BEING VIEWED BY OTHERS LINKAGE MAPPING AND QTL

ANALYSIS OF FLOWERING TIME IN FABA BEAN Article Open access 02 July 2021 GRAS-DI SYSTEM FACILITATES HIGH-DENSITY GENETIC MAP CONSTRUCTION AND QTL IDENTIFICATION IN RECOMBINANT INBRED LINES

OF THE WHEAT PROGENITOR _AEGILOPS TAUSCHII_ Article Open access 08 December 2020 RECOMBINANT INBRED LINES AND NEXT-GENERATION SEQUENCING ENABLE RAPID IDENTIFICATION OF CANDIDATE GENES

INVOLVED IN MORPHOLOGICAL AND AGRONOMIC TRAITS IN FOXTAIL MILLET Article Open access 07 January 2022 INTRODUCTION Adzuki bean (_Vigna angularis_ (Willd.) Ohwi & Ohashi), Xiaodou (red

beans) in Chinese, belongs to the subgenus _Ceratotropis_ of the _Vigna_ genus in the Phaseoleae tribe. It is a self-pollinating diploid plant with 2n = 2x = 22 chromosomes1. Adzuki bean is

one of the most economically important legume food crops in Asia and is primarily cultivated in China, Japan, and South Korea, as well as other countries2. The protein content of adzuki bean

is two to three times that of cereal crops. Recently, adzuki bean has been recommended as a suitable food for diabetic patients due to its excellent protein and phenolic compound

contents3,4. In China, adzuki bean is a traditional healthy food, and the market demand for adzuki bean products has increased gradually as consumers have become more health conscious.

However, unlike other major crops (such as rice, maize, and wheat), progress in adzuki bean genetics and breeding is far from satisfactory5. The construction of a high-quality genetic

linkage map will enable the discovery of useful genes and accelerate the breeding of adzuki bean. Several adzuki bean genetic linkage maps have been constructed using different F2 or BC1F1

mapping populations6,7,8,9,10. Among these maps, the most comprehensive genetic linkage map consists of 11 linkage groups (LGs) spanning a total of 832.1 cM, with an average distance of 1.85

 cM between markers, including 205 simple sequence repeat(SSR) markers, 187 amplified fragment length polymorphism(AFLP) markers, and 94 restriction fragment length polymorphism (RFLP)

markers8. Quantitative trait loci (QTLs) associated with several important traits have been identified using this linkage map, including those related to seeds, pods, stems and leaves11, as

well as bruchid resistance12. The usefulness of genetic maps largely depends on their density13: a high-density linkage map will promote high-resolution genetic mapping and positional

cloning of crucial genes and can also benefit physical map assembly. To construct a high-density genetic map of adzuki bean, additional molecular markers need to be developed.

Single-nucleotide polymorphisms (SNPs) are the most abundant class of polymorphisms in most genomes and are one of the most efficient markers for identifying candidate genes associated with

QTLs14. Based on the development of next-generation sequencing technology, several high-throughput methods for SNP and insertion/deletion polymorphism (InDel) marker identification and

genotyping have been developed. These methods include restriction site-associated (RAD) sequencing (RADseq)15, genotyping-by-sequencing(GBS)16, and specific length amplified fragment

sequencing (SLAF-seq)17. Among these methods, SLAF-seq combines pre-designed reduced representation library (RRL) schemes, high-throughput paired-end sequencing technology, and a double

barcode system, which allows it to simultaneously genotype large populations with a considerable number of loci at a lower cost. Importantly, reference genome sequences and polymorphism

information are not necessary when this method is used. This method has been applied in many species for genetic map construction as a rapid and cost-effective strategy for high-throughput

SNP and InDel discovery and genotyping18,19,20. Flowering time is a very important target in adzuki bean breeding programmes because it is critical for adapting cultivars to different

cultivation areas or growing seasons. A series of genes or QTLs related to flowering time have been detected in Arabidopsis21,22, rice23, wheat24, soybean25,26, common bean27,28, and other

plants. However, few studies have focused on candidate genes or QTLs for flowering time in adzuki bean. Two studies that focused on the genetics of domestication in adzuki bean described 1

to 5 QTLs related to first flowering time (FLD)9,11. For example, Isemura _et al_. detected a QTL for FLD (_Fld2.4.1,_ phenotypic varianceexplained (PVE): 23.9%) on LG 4 using an F2

population11. Kaga _et al_. identified one major QTL of FLD (_Fld3.4a.1,_ PVE: 43.7%) on LG 4a and four QTLs with smaller effects on LGs2 (_Fld3.2.1,_ PVE: 6%), 3(_Fld3.3.1_, PVE: _5.4%_),

5(_Fld3.5.1_, PVE: _8.8%_), and 11(_Fld3.11.1_, PVE:_5.8%_) also using an F2 population9. However, the precise genomic positions of these QTLs remain unclear, and the genes underlying these

flowering time QTLs in adzuki bean are not known. The main objective of this research was to analyse a recombinant inbred line(RIL) mapping population from a cross between wild and

cultivated adzuki bean for SNP and InDel polymorphisms and for QTLs associated with flowering time during different years. The specific objectives were (a) to construct a high-density

genetic map of adzuki bean based on the SLAF-seq high-throughput method and (b) to identify QTLs associated with FLD over two years. RESULTS SLAF SEQUENCING AND GENOTYPING SLAF library

construction and Illumina sequencing generated a total of 47.8 Gb of raw data containing 240,904,338 paired-end reads with a length of 100 bp. The Q20 ratio (a quality score of 20,

indicating a 1% chance of an error) was 81.12%, and the GC (guanine-cytosine) content is 33%. Of the high-quality data, 45.3 M reads and 71.1 M reads were obtained from the male and female

parents, respectively. After clustering the reads, a total of 243,980 SLAF loci were detected. The average sequence depths of the SLAF loci were 190.14-fold for the male parent, 122.63-fold

for the female parent, with an average of 4.25-fold for each individual offspring. A total of 23,294 polymorphic SLAF markers were detected between the two parents, of which 21,421 were

successfully encoded and grouped into eight segregation patterns(i.e., ab × cd, ef × eg, hk × hk, lm × ll, nn × np, aa × bb, ab × cc and cc × ab). Because only the aa × bb segregation

pattern was suitable for the RIL population, the 13,158 SLAFs in this segregation pattern were selected for further analysis. After discarding low-quality SLAFs with a < 10-fold parental

sequencing depth and <70% integrity among the RIL population, 10,890 high-quality SLAFs remained for use in the linkage analysis. GENETIC LINKAGE MAP CONSTRUCTION After discarding SLAF

markers with a pair-wise independence LOD < 5, a total of 4425 markers could be assigned to 11 LGs. The coverage of the markers was 379.19-fold for the male parent, 243.04-fold for the

female parent, with an average of 4.46-fold for each RIL individual. Of the assigned SLAF markers, 1938 markers did not segregate in the expected 1:1 ratio, as based on a chi-square test

where the P-value of a marker segregating1:1 was <1%. The markers with distorted segregation were initially excluded from the map construction but were added later as accessory markers.

To more reasonably calculate the average distance between adjacent markers, 3253 markers with the same genotype across the entire RIL population were merged into 861 bins. The bin

information is supplied in Supplementary Table S1. The final map is 1628.15 cM in length with an average marker density of 0.80 cM (Fig. 1 and Table 1). The number of SLAF markers in each LG

ranges from 108 (LG06) to 622 (LG07). The length of each LG ranges from 61.05 cM (LG06) to 270.83 cM (LG07), and the average distance between adjacent markers is 0.61 cM (LG05) to 0.95 cM

(LG09). Except for one >6-cM gap between adjacent markers on LG09, all other gaps are <4.1 cM. COMPARATIVE ANALYSIS To evaluate the quality of the genetic map, the sequences of the

4425 mapped SLAF markers were aligned to the two draft genome sequences of adzuki bean using the stringent threshold described in the materials and methods section. Based on the draft adzuki

bean genome reported by Kang _et al_.29, the BLAST result indicated that 76.14% (3369/4425) of the SLAF markers map to pseudo-chromosomes or scaffolds (Supplementary Table S2). The SLAF

markers on LG01, LG04, LG09, LG10, and LG11 primarily map to a single pseudo-chromosome (Table 2). The high Spearman’s rank correlation coefficient (>0.9) indicated high collinearity

between the LGs and pseudo-chromosomes. The SLAF markers on LG02, LG03, LG05, LG06, LG07, and LG08 predominantly map to two to four pseudo-chromosomes or long scaffolds. A similar analysis

was performed for the adzuki bean genome reported by Yang _et al_.30. The BLAST results showed that 75.16% (3326/4425) of the SLAF markers map to pseudo-chromosomes or scaffolds

(Supplementary Table S3). Except for LG01, LG03, LG05 and LG09, all other LGs are primarily homologous with a single pseudo-chromosome, with high synteny (Table 2). To align the genetic map

with the reference genome of common bean31, macrosynteny and microsyntenywere detected between adzuki bean and common bean. All 11 LGs are syntenic with 10 common bean chromosomes (Fig. 2

and Supplementary Table S4). LG01, LG03, LG04, LG06 and LG09 correspond to only one common bean chromosome. The locus order of LG09 is most similar to common bean chromosome 6 (Pchr6). LG02,

LG05, LG07, LG08, LG10 and LG11 mainly correspond to two common bean chromosomes. Interestingly, LG02 and LG07 are syntenic with common bean chromosomes 2 (Pchr2) and 3 (Pchr3) according to

a ‘sharing’ model, which suggests that these two chromosomes in mungbean and common bean have recombined. QTLS FOR FIRST FLOWERING TIME FLD was evaluated over two years. As expected, the

FLD of the wild parent (Yesheng10) occurred much later than did that of the cultivated parent (Jihong9218), at 78 days compared to 43 days. Transgressive segregation lines were observed in

the RIL population. The distribution of FLD among RIL population was nearly binomial (Fig. 3). A total of 8 QTLs were detected for FLD during the 2 years of the study (Fig. 4 and Table 3).

Four QTLs were identified each year. Among them, two major QTLs (_Fld3.2_ and _Fld3.3_) for FLD with high LOD values (39.17 in 2013 and 35.63 in 2014) were consistently identified at the

same map position as LG03 during both years, and the PVE by the QTLs was 70.9% (2013) and 66% (2014). Another 4 minor QTLs (_Fld5.1_ vs _Fld5.3,_ and _Fld5.2_ vs _Fld5.5_) for FLD were

detected at a similar map position on LG05 during both years. However, _Fld3.1_ on LG03 and _Fld5.4_ on LG05 were identified only during one year. The alleles of all QTLs from the wild

parent delayed flowering time. DISCUSSION The genetic map developed in this study contains 4425 SLAF markers, a majority of which are anchored to the adzuki bean draft genome scaffolds. The

map consists of 11 LGs corresponding to the haploid chromosome number of the _Vigna_ genus. Compared to the genetic map constructed by Han _et al_.8, the number of mapped loci (486 vs 4425),

marker density (1.85 cM vs 0.8 cM), and total map length (832.1 cM vs 1628.15 cM) are significantly improved in this dense genetic linkage map. For our high-density genetic linkage map, the

maximum genetic distance between flanking markers is 6.63 cM, which is much less than the 18.5 cM for the map of Han _et al_. Compared with the two SNP maps constructed by Kang _et al_.29

and Yang _et al_.30 to assemble the adzuki bean draft genome, the marker number in this individual map is substantially higher. However, the proportion of markers showing segregation

distortion (43.8%) is higher than for the previous interspecific mapping populations of _V. nepalensis_ x _V. angularis_(3.9%)8, _V. angularis_ x _V. nakashimae_ (19.7%)6 and _V. umbellata_

x _V. angularis_ (29.8%)7. One reason for these differences in segregation distortion may be the use of an RIL mapping population; in contrast, the above mentioned interspecific mapping

populations were BC1F1 and F2 populations. The skewed segregation ratio of RIL populations is usually higher than that of BC and F2 populations. A similar phenomenon was observed during the

construction of maps for rice32 and chickpea33. Another possible reason maybe that from generations F2 to F8 during RIL population generation, 19 lines were lost due to disease. However,

compared to using BC or F2 as mapping populations, using an RIL mapping population can effectively reduce the influence of the dominant effect and reveal the additive effect of QTLs. Most

importantly, an RIL mapping population as a permanent population can be planted in different years or environments, which would improve the accuracy of QTL mapping. Because discarding the

markers with segregated distortion may not only remove biologically interesting segments of the genome but also dramatically reduce the marker density of the genetic map and because markers

with skewed segregation can be successfully used for QTL mapping34, we chose to include the markers with distorted segregation as accessory markers after map construction. Markers with the

same genotype across the mapping population should be binned when constructing a high-density map; otherwise, the average inter-marker distance may not reflect the real distribution of the

markers in the genome. A series of dense maps have been constructed according to this principle, such as the ultra-dense genetic maps of potato35, rice36, sorghum37, and maize38. In this

study, the average inter-marker distance was 0.37 cM before we binned the SLAF markers from RIL with the same genotype but 0.80 cM after merging. Although this principle doubled the average

inter-marker distance, it was obviously more reasonable. A comparison of the dense genetic linkage map and two physical maps showed that the marker order along chromosomes was mostly

collinear (Table 2), which confirmed the high quality of our map. However, we also found that several LGs mapped to more than one chromosome. For example, LG02, LG03, LG05, LG06, LG07, and

LG08 mapped to 3, 2, 3, 3, 4, and 3 chromosomes, respectively, when aligned with the physical map of Kang _et al_.29. Four LGs (LG01, LG03, LG05 and LG09) mapped to 2 chromosomes in the

physical map of Yang _et al_.30. To investigate these contradicting results and to determine the relationship between our map and that of Han _et al_.8, we also aligned 186 pairs of SSR

primer sequences from the map of Han _et al_. with the two reference genomes, and similar results were obtained (Supplementary Table S5). For example, LG 1 of Han’s map mainly mapped to

chromosomes 10 and 11 and scaffold 33 when aligned with the physical map by Kang _et al_.29, and it primarily mapped to chromosomes 1, 7, and 11 when aligned with the physical map of Yang

_et al_.30. The mapping of LG07 from our high-density genetic linkage map produced similar results. Indeed, the results from mapping our genetic map as well as that of Han _et al_. to the

two physical maps are in strong agreement for all 11 LGs, which will facilitate comparisons among these LGs. The mapping of LGs to more than one chromosome may be explained by genotyping

errors, misassembled scaffolds or reciprocal translocations between chromosomes. We suggest that the mapping of LG05 to chromosomes 3 and 8 on the physical map of Yang _et al_.30 is due to a

reciprocal translocation because it corresponds to an “LG 4 + 6” reciprocal translocation model in some wild adzuki bean accessions39. However, further efforts are needed to confirm this

inference. Based on the number of SLAF markers mapped to pseudo-chromosomes and the collinearity between LGs and pseudo-chromosomes, we determined that the assembly of the adzuki bean draft

genome reported by Yang _et al_.30 is more accurate than that reported by Kang _et al_.29. This is in agreement with the genome assembly coverage (79.9% vs 75% of adzuki bean genome) results

of the two physical maps. However, the assembly of pseudo-chromosome 1 on the physical map of Kang _et al_. may be better than the corresponding pseudo-chromosome on that of Yang _et al_.,

as it has more identified SLAF markers and better collinearity. This conclusion may be helpful when selecting a suitable reference genome for fine-mapping genes. Moreover, a high-resolution

linkage map is useful for physical map assembly using next-generation sequencing. The mapping of SLAF markers to scaffolds will improve the ordering and orientation of the remaining unplaced

sequences in the two draft genome sequence databases of adzuki bean. Genomic synteny analysis is useful for comparative genomics. Given the close genetic relationship between common bean

and adzuki bean, we aligned the genetic map with the reference genome of common bean. Our results demonstrated high synteny between the two genomes (Fig. 2), which suggests that candidate

genes could be identified through comparative mapping. In addition, the extensive synteny and collinearity observed between the common bean genome and the current map provide additional

support for the mapping accuracy of our high-density genetic map. Flowering time is critical for adapting adzuki bean cultivars to different cultivation areas or growing seasons. Isemura _et

al_. detected a QTL (_Fld2.4.1_) of FLD on LG 4 (map of Han _et al_.) after phenotyping an F2 population derived from a cross between cultivated adzuki bean (_Vigna angularis_) and wild

relative (_Vigna nepalensis_) in a single year11. Kaga _et al_. identified one major QTL of FLD (_Fld3.4a.1_) on LG 4a and four QTLs with smaller effects on LGs2 (_Fld3.2.1_), 3(_Fld3.3.1_),

5(_Fld3.5.1_), and 11(_Fld3.11.1_) using an F2 population derived from a cross between cultivated adzuki bean (_Vigna angularis_) and wild adzuki bean (_Vigna nipponensis_) accession9. In

the present study, a two-year phenotypic evaluation of flowering time in an RIL population produced from an early-flowering adzuki bean cultivar and a late-flowering wild adzuki bean

accession indicated the presence of one major QTL on LG03 and two minor QTLs on LG05. Because both studies mentioned above were based on the genetic map of Han _et al_. and we had aligned

both our map and that of Han _et al_. to the two physical maps of adzuki bean (Supplementary Table S5), it is possible to compare the QTLs from all three studies. Based on the alignment with

the physical map of Yang _et al_., _Fld2.4.1_ (the nearest SSR markers were CEDG103 and CEDG011) is located on chromosome 3, and its physical position is between 32.8 and 39.1 Mb.

Similarly, _Fld3.4a.1_ (the nearest SSR markers were CEDG036 and CEDG127) is also located on chromosome 3, and the physical positionis between 36.2 and 40.0 Mb. Because these two QTLs are

located in the same genomic region, they may be influenced by the same gene or genes. _Fld3.2.1_ maps to chromosome 4 in the physical map of Yang _et al_.; in this study, _Fld3.2_ and

_Fld3.3_ were detected on the same chromosome. A neighbouring SSR marker (CEDG026) showed that its physical position is approximately 1.5 Mb on chromosome 4, and the distance from the

associated marker (Marker56693) of _Fld3.2_ and _Fld3.3_ is approximately 2.2 Mb, indicating that they may be located at the same QTL. However, similar to _Fld2.4.1_(PVE: 23.9%) vs

_Fld3.4a.1_(PVE: 43.7%), the PVEs by _Fld3.2.1_ (PVE: 6.0%) and _Fld3.2_(PVE: 70.9%) differed greatly. These differences may be due to allelic variation or interactions between the QTL and

the genetic background. Further research is needed to verify this inference. Considering that LG03 shows high synteny with chromosome 9 (Pchr9) of the common bean genome (Fig. 2), we

assessed QTLs for flowering time on this chromosome. As expected, we found flowering time-related genes (i.e., _PvZTL_) on LGB9 (corresponding to Pchr9 in the common bean reference

genome)28, which are located near the GH locus identified by Tar’an _et al_.40. Using the ZTL protein sequence of _Arabidopsis_, the adzuki bean homologues in GenBank were identified by a

BLAST search as located near 0.2 Mb on chromosome 4; the distance from the associated marker (Marker56693) of _Fld3.2_ and _Fld3.3_ is approximately 3.5 Mb. Further studies are needed to

confirm the role of this homologue as a potential flowering time gene. In summary, this study is the first attempt to conduct QTL analysis using an NGS-derived dense genetic map in adzuki

bean. The results provide a foundation that will be useful for future genomic research, such as the identification of QTLs for other important traits, positional cloning, comparative mapping

in legumes, ordering and orienting the remaining unplaced scaffolds in the two draft genomes of adzuki bean and marker-assisted selection in adzuki bean breeding. MATERIALS AND METHODS

MAPPING POPULATION An F8:9RIL population was developed from the cross between a wild adzuki bean accession (_Vigna nipponensis_: Yesheng10) collected in Dandong, China(39.75N, 123.74E), and

an adzuki bean cultivar (Jihong9218) that is widely grown in northern China. The cultivated parent had an early flowering time and was the male in the cross. The wild adzuki bean accession

had a late flowering time and was the female in the cross. The RIL population consisted of 153 lines generated from a single seed descent from generations F2 to F8. The parental accessions

used in the cross were obtained from the Hebei Academy of Agricultural and Forestry Sciences (HAAFS). PHENOTYPIC EVALUATION FOR VARIATION IN FLOWERING TIME The RIL population of 153 lines

and 20 plants of each parent was grown in the field at HAAFS, Shijiazhuang, China (37.95N, 114.73E), from June to November in 2013 and 2014. The soil at the site was a sandy loam with no

major fertility problems (PH = 7.3). Experimental units consisted of two-row plots 3.5 m long and 1 m wide. Each planting was a randomized block design experiment that was repeated twice.

The FLD (days from sowing to the first flower) was evaluated using the mean value of each RIL line and parent over a two-year period (2013 and 2014). DNA EXTRACTION Young leaves from 153 RIL

lines and two parents were collected, immediately frozen in liquid nitrogen, and transferred to a −80 °C freezer. To obtain high-quality DNA for SLAF library construction, a plant DNAzol

kit (ThermoFisher, Waltham, MA, USA) was used to extract the total genomic DNA according to the manufacturer’s instructions. SLAF LIBRARY CONSTRUCTION AND HIGH-THROUGHPUT SEQUENCING The

procedure used for SLAF library construction was conducted as described by Sun _et al_.17 with minor modifications. In brief, a draft reference genome of adzuki bean29 was used to design

SLAF marker discovery experiments by simulating in silico the number of markers produced by restriction digest with two different enzyme combinations. Accordingly, an SLAF pilot experiment

was performed, and the SLAF libraries were constructed. Two enzymes (HaeIII and SspI-HF; NEB, Ipswich, MA, USA) were used to digest the genomic DNA of the parents and RIL population.

Subsequently, a single nucleotide (A) overhang was added to the digested fragments using Klenow Fragment enzyme (NEB, Ipswich, MA, USA) and ATP at 37 °C. Next, duplex tag-labelled sequencing

adapters (Life Technologies, Carlsbad, California, USA) were ligated to the A-tailed fragments using T4 DNA ligase (NEB, Ipswich, MA, USA). Then, the diluted and ligated DNA samples were

used as a template for polymerase chain reaction (PCR). Each PCR reaction also contained High-Fidelity DNA Polymerase (NEB, Ipswich, MA, USA), dNTPs and PCR primers (Forward sequence:

5′-AATGATACGGCGACCACCGA-3′, reverse sequence: 5′-CAAGCAGAAGACGGCATACG-3′) (PAGE-purified, Life Technologies). PCR products were purified using Quick Spin columns (Qiagen, Venlo, Netherlands)

and separated by 2% agarose gel electrophoresis. SLAFs of 264–414 bp (with adapter sequence indexes and adapters) in size were excised and purified using a QIAquick Gel Extraction kit

(Qiagen, Hilden, Germany). Finally, 100-bp paired-end sequencing was performed on an Illumina HiSeq 2500 sequencing platform (Illumina, Inc., San Diego, CA, USA) according to the

manufacturer’s instructions. SEQUENCE DATA GROUPING AND GENOTYPE DEFINITION The procedures used for SLAF marker identification and genotyping were performed as described by Sun _et al_.17.

Briefly, low-quality reads (quality score < 20) were filtered out, and the barcodes were trimmed from each high-quality read. All of the clean reads were clustered based on sequence

similarity as determined by BLAST (−tileSize = 10, −stepSize = 5). Sequences with >95% identity were grouped into one SLAF locus. Allele tags of each SLAF locus with a sequencing depth

>10-fold for parental reads and >70% integrity in the offspring were collected. Both SNP and InDel loci were detected between parents, and SLAFs with >3 SNPs or InDels were filtered

out. Because adzuki bean is adiploid species and one locus contains at most four SNP tags, groups containing more than four tags were discarded. Only SLAFs with 2, 3, or 4 tags were

identified as polymorphic and considered potential markers. All polymorphic markers were classified into eight segregation patterns (aa × bb, ab × cd, ef × eg, hk × hk, lm × ll, nn × np, ab 

× cc and cc × ab). Because the mapping population was composed of RILs, only the aa × bb segregation pattern was used for genetic linkage map construction. GENETIC LINKAGE MAP CONSTRUCTION

JoinMap ver. 4.041 was used to construct a linkage map. Marker segregation ratios were calculated using the chi-square test. Markers showing significant (P < 0.01) segregation distortion

were initially excluded from the map construction but were added later as accessory markers. Markers with the same genotype across the entire RIL population were binned35,38. The grouping

and ordering of the markers were established using a maximum likelihood algorithm. Pairwise marker loci that showed a likelihood-ratio statistic (LOD) value larger than 5.0 were used to

create LGs, and the recombination frequencies were converted into map distances (cM) using the Kosambi mapping function42. COMPARATIVE ANALYSIS OF THE HIGH-DENSITY LINKAGE MAP Colinearity

between the high-density linkage map and the adzuki bean genome was determined by comparing the assigned SLAF sequences to two adzuki bean draft genomes29,30 with BLASTN (version 2.2.30) and

a cut-off value of e−30. SLAF markers with query sequence lengths ≥80 and sequence identity >98% were selected to calculate the Spearman rank correlation according to their order on LGs

and physical position on chromosomes using R (version 3.1.2) software. Synteny between adzuki bean and common bean (_Phaseolus vulgaris_) was determined with BLASTN searches against the

genome of _Phaseolus vulgaris_ (http://phytozome.jgi.doe.gov/pz/portal.html) using the source sequences of mapped SLAF markers as queries. The significance cut-off value was e−20 for an

overlap of at least 70 bp. Synteny was visualized with MapChart (version 2.3) software43. QTL ANALYSIS QTL analysis was conducted with MapQTL ver.6 as described by Van _et al_.44. Briefly,

the entire genome was scanned for QTLs of FLD using a general interval mapping (IM) method. A regression algorithm was used to calculate the maximum likelihood, and a 1.0 mapping step size

was used. The significance of each QTL interval was tested with an LOD. The threshold of the LOD score for significance (P = 0.05) was determined using 10,000 permutations. The PVE by each

QTL was estimated based on the population variance found within the segregating population. ADDITIONAL INFORMATION HOW TO CITE THIS ARTICLE: Liu, C. _et al_. Development of a high-density

genetic linkage map and identification of flowering time QTLs in adzuki bean(_Vigna angularis_). _Sci. Rep._ 6, 39523; doi: 10.1038/srep39523 (2016). PUBLISHER'S NOTE: Springer Nature

remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. REFERENCES * Zheng, Z. Adzuki bean In Food Legumes in China (ed. Zheng, Z. ) 173–195

(China Agriculture Press, 1997). * Vaughan, D. A., Tomooka, N. & Kaga, A. Azuki bean [Vigna angularis (L.) Ohwi and Ohashi]. In Genetic resources, chromosome engineering, and crop

improvement. Grain legumes. Volume I (ed. Singh, R. J. & Jauhar, P. P. ) 341–353 (CRC Press, 2005). * Lin, P. Y. & Lai, H. M. Bioactive compounds in legumes and their germinated

products. Journal of agricultural and food chemistry 54, 3807–3814 (2006). Article  CAS  PubMed  Google Scholar  * Yao, Y., Cheng, X., Wang, L., Wang, S. & Ren, G. A determination of

potential alpha-glucosidase inhibitors from Azuki Beans (_Vigna angularis_). International journal of molecular sciences 12, 6445–6451 (2011). Article  CAS  PubMed  PubMed Central  Google

Scholar  * Cheng, X. Z. & Tian, J. Status and future perspectives of Vigna (mungbean and azuki bean) production and research in China In The 14th NIAS international workshop on genetic

resources–Genetic resources and comparative genomics of legumes (Glycine and Vigna) (ed. Tomooka, N. & Vaughan, D. A. ), Tsukuba: National Institute of Agrobiological Science, 83–86

(2011). * Kaga, A., Ohnishi, M., Ishii, T. & Kamijima, O. A genetic linkage map of azuki bean constructed with molecular and morphological markers using an interspecific population

(_Vigna angularis_ x _V. nakashimae_). Theoretical & Applied Genetics 93, 658–663 (1996). Article  CAS  Google Scholar  * Kaga, A., Ishii, T., Tsukimoto, K., Tokoro, E. & Kamijima,

O. Comparative molecular mapping in Ceratotropis species using an interspecific cross between azuki bean (_Vigna angularis_) and rice bean (_V. umbellata_). Theoretical & Applied

Genetics 100, 207–213 (1999). Article  Google Scholar  * Han, O. K. et al. A genetic linkage map for azuki bean [_Vigna angularis_ (Willd.) Ohwi & Ohashi]. Theoretical and applied

genetics 111, 1278–1287 (2005). Article  CAS  PubMed  Google Scholar  * Kaga, A., Isemura, T., Tomooka, N. & Vaughan, D. A. The genetics of domestication of the azuki bean (_Vigna

angularis_). Genetics 178, 1013–1036 (2008). Article  CAS  PubMed  PubMed Central  Google Scholar  * Luo, W. X. et al. Construction of Genetic Linkage Map Using SSR Molecular Markers in

Azuki Bean (_Vigna angularis_ Ohwi and Ohashi). _Scientia_ Agricultura Sinica 46, 3534–3544 (2013). CAS  Google Scholar  * Isemura, T. et al. Genome dissection of traits related to

domestication in azuki bean (_Vigna angularis_) and comparison with other warm-season legumes. Ann Bot 100, 1053–1071 (2007). Article  PubMed  PubMed Central  Google Scholar  * Somta, P. et

al. Mapping of quantitative trait loci for a new source of resistance to bruchids in the wild species Vigna nepalensis Tateishi & Maxted (_Vigna_ subgenus _Ceratotropis_). Theoretical

and applied genetics 117, 621–628 (2008). Article  CAS  PubMed  Google Scholar  * Harushima, Y. et al. A High-Density Rice Genetic Linkage Map with 2275 Markers Using a Single F2 Population.

Genetics 1998, 479–494 (1998). Google Scholar  * Rafalski, A. Applications of single nucleotide polymorphisms in crop genetics. Current Opinion in Plant Biology 5, 94–100 (2002). Article 

CAS  PubMed  Google Scholar  * Davey, J. W., Davey, J. L., Blaxter, M. L. & Blaxter, M. W. RADSeq: next-generation population genetics. Briefings in Functional Genomics 9, 108 (2011).

Google Scholar  * Elshire, R. J. et al. A Robust, Simple Genotyping-by-Sequencing (GBS) Approach for High Diversity Species. PloS one 6, e19379 (2011). Article  CAS  ADS  PubMed  PubMed

Central  Google Scholar  * Sun, X. et al. SLAF-seq: An Efficient Method of Large-Scale _de novo_ SNP Discovery and Genotyping Using High-Throughput Sequencing. PloS one 8, e58700 (2013).

Article  CAS  ADS  PubMed  PubMed Central  Google Scholar  * Zhang, Y. et al. Construction of a high-density genetic map for sesame based on large scale marker development by specific length

amplified fragment (SLAF) sequencing. BMC plant biology 13, 141 (2013). Article  ADS  PubMed  PubMed Central  Google Scholar  * Zhang, J. et al. High-density genetic map construction and

identification of a locus controlling weeping trait in an ornamental woody plant (Prunus mume Sieb. et Zucc). DNA research: an international journal for rapid publication of reports on genes

& genomes 22, 183 (2015). Article  CAS  Google Scholar  * Qi, Z. et al. A high-density genetic map for soybean based on specific length amplified fragment sequencing. PloS one 9,

e104871 (2014). Article  ADS  PubMed  PubMed Central  Google Scholar  * Somers, D. E., Schultz, T. F., Milnamow, M. & Kay, S. A. ZEITLUPE encodes a novel clock-associated PAS protein

from _Arabidopsis_. Cell 101, 319–329 (2000). Article  CAS  PubMed  Google Scholar  * Pedrero, M., Campuzano, S. & Pingarrón, J. M. CONSTANS mediates between the circadian clock and the

control of flowering in _Arabidopsis_. Nature 410, 1116–1120 (2001). Article  Google Scholar  * Zhao, X. L., Shi, Z. Y., Peng, L. T., Shen, G. Z. & Zhang, J. L. An atypical HLH protein

OsLF in rice regulates flowering time and interacts with OsPIL13 and OsPIL15. New Biotechnology 28, 788–797 (2011). Article  CAS  PubMed  Google Scholar  * Nemoto, Y., Kisaka, M., Fuse, T.,

Yano, M. & Ogihara, Y. Characterization and functional analysis of three wheat genes with homology to the CONSTANS flowering time gene in transgenic rice. Plant Journal 36, 82–93 (2003).

Article  CAS  Google Scholar  * Tasma, I. M. & Shoemaker, R. C. Mapping Flowering Time Gene Homologs in Soybean and Their Association with Maturity (_E_) Loci. Crop Science 43, 319–328

(2003). Article  CAS  Google Scholar  * Yamanaka, N. et al. Fine mapping of the FT1 locus for soybean flowering time using a residual heterozygous line derived from a recombinant inbred

line. Theoretical and applied genetics 110, 634–639 (2005). Article  CAS  PubMed  Google Scholar  * Koinange, E. M. K., Singh, S. P. & Gepts, P. Genetic Control of the Domestication

Syndrome in Common Bean. Crop Science 36, 1037–1045(1996). Article  Google Scholar  * Kwak, M., Velasco, D. & Gepts, P. Mapping homologous sequences for determinacy and photoperiod

sensitivity in common bean (_Phaseolus vulgaris_). The Journal of heredity 99, 283–291 (2008). Article  CAS  PubMed  Google Scholar  * Kang, Y. J. et al. Draft genome sequence of adzuki

bean, _Vigna angularis_. Scientific reports 5, 8069 (2015). Article  PubMed  PubMed Central  Google Scholar  * Yang, K. et al. Genome sequencing of adzuki bean (_Vigna angularis_) provides

insight into high starch and low fat accumulation and domestication. Proceedings of the National Academy of Sciences of the United States of America 112, 13213–8 (2015). Article  CAS  ADS 

PubMed  PubMed Central  Google Scholar  * Kiviluoto, Luukkonen, Salo & Kivilaakso . A reference genome for common bean and genome-wide analysis of dual domestications. Nature genetics

46, 707–713 (2014). Article  Google Scholar  * Liang, Y. S. et al. Comparison of genetic linkage maps based on F2 & F6 populations derived from rice subspecies cross. Hereditas 29,

1110–1120 (2007). Article  CAS  PubMed  Google Scholar  * Gaur, R. et al. High-throughput SNP discovery and genotyping for constructing a saturated linkage map of chickpea (_Cicer arietinum_

L.). DNA research: an international journal for rapid publication of reports on genes and genomes 19, 357–373 (2012). Article  CAS  Google Scholar  * Xu, S. Quantitative trait locus mapping

can benefit from segregation distortion. Genetics 180, 2201–2208 (2008). Article  PubMed  PubMed Central  Google Scholar  * van Os, H. et al. Construction of a 10,000-marker ultradense

genetic recombination map of potato: providing a framework for accelerated gene isolation and a genomewide physical map. Genetics 173, 1075–1087 (2006). Article  CAS  PubMed  PubMed Central

  Google Scholar  * Xie, W. et al. Parent-independent genotyping for constructing an ultrahigh-density linkage map based on population sequencing. Proceedings of the National Academy of

Sciences of the United States of America 107, 10578–10583 (2010). Article  CAS  ADS  PubMed  PubMed Central  Google Scholar  * Zou, G. et al. Identification of QTLs for eight agronomically

important traits using an ultra-high-density map based on SNPs generated from high-throughput sequencing in sorghum under contrasting photoperiods. Journal of experimental botany 63,

5451–5462 (2012). Article  CAS  PubMed  Google Scholar  * Guo, T. et al. Genetic basis of grain yield heterosis in an “immortalized F(2)” maize population. Theoretical and applied genetics

127, 2149–2158 (2014). Article  PubMed  Google Scholar  * Wang, L. et al. Reciprocal translocation identified in _Vigna angularis_ dominates the wild population in East Japan. Journal of

plant research 128, 1–8 (2015). Article  Google Scholar  * Tar’an, B., Michaels, T. E. & Pauls, K. P. Genetic Mapping of Agronomic Traits in Common Bean. Crop Science 42, 544–556 (2002).

Article  Google Scholar  * Van Ooijen, J. W. JoinMap 4: Software for the calculation of genetic linkage maps in experimental populations, Wageningen, Netherlands. URL

https://www.kyazma.nl/index.php/JoinMap (2006). * Kosambi, D. D. The estimation of map distances from recombination values. Annals of Human Genetics 12, 172–175 (1943). Google Scholar  *

Voorrips, R. E. MapChart: Software for the Graphical Presentation of Linkage Maps and QTLs, Wageningen, Netherlands. URL https://www.wur.nl/en/show/Mapchart-2.30.htm (2002). * Van Ooijen, J.

W. MapQTL 6: software for the mapping of quantitative trait loci in experimental populations of diploid species. Wageningen, Netherlands. URL https://www.kyazma.nl/index.php/mc.MapQTL

(2009). Download references ACKNOWLEDGEMENTS This research was supported by China Agriculture Research System (CARS-09), National Science and Technology Support Project (2014BAD07B05), and

Science and Technology Support Program of Hebei Province (16227508D). AUTHOR INFORMATION AUTHORS AND AFFILIATIONS * Key Laboratory of Crop Genetics and Breeding of Hebei Province; Institute

of Cereal and Oil Crops, Hebei Academy of Agricultural and Forestry Sciences, Shijiazhuang, 050035, China Changyou Liu, Baojie Fan, Zhimin Cao, Qiuzhu Su, Yan Wang, Zhixiao Zhang & Jing

Tian Authors * Changyou Liu View author publications You can also search for this author inPubMed Google Scholar * Baojie Fan View author publications You can also search for this author

inPubMed Google Scholar * Zhimin Cao View author publications You can also search for this author inPubMed Google Scholar * Qiuzhu Su View author publications You can also search for this

author inPubMed Google Scholar * Yan Wang View author publications You can also search for this author inPubMed Google Scholar * Zhixiao Zhang View author publications You can also search

for this author inPubMed Google Scholar * Jing Tian View author publications You can also search for this author inPubMed Google Scholar CONTRIBUTIONS Jing Tian and Changyou Liu designed the

experiment. Changyou Liu carried out the DNA isolation and sequence data analyses and drafted the manuscript. Jing Tian assisted in manuscript preparation. Baojie Fan, Zhimin Cao, Qiuzhu

Su, Yan Wang, and Zhixiao Zhang prepared the plant materials and participated in DNA extraction. All of the authors read and approved the final manuscript. ETHICS DECLARATIONS COMPETING

INTERESTS The authors declare no competing financial interests. ELECTRONIC SUPPLEMENTARY MATERIAL SUPPLEMENTARY INFORMATION RIGHTS AND PERMISSIONS This work is licensed under a Creative

Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in

the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of

this license, visit http://creativecommons.org/licenses/by/4.0/ Reprints and permissions ABOUT THIS ARTICLE CITE THIS ARTICLE Liu, C., Fan, B., Cao, Z. _et al._ Development of a

high-density genetic linkage map and identification of flowering time QTLs in adzuki bean (_Vigna angularis_). _Sci Rep_ 6, 39523 (2016). https://doi.org/10.1038/srep39523 Download citation

* Received: 18 August 2016 * Accepted: 23 November 2016 * Published: 23 December 2016 * DOI: https://doi.org/10.1038/srep39523 SHARE THIS ARTICLE Anyone you share the following link with

will be able to read this content: Get shareable link Sorry, a shareable link is not currently available for this article. Copy to clipboard Provided by the Springer Nature SharedIt

content-sharing initiative