https://orcid.org/0000-0003-3355-4803
Figures
Abstract
Estimation of genetic diversity in rapeseed is important for sustainable breeding program to provide an option for the development of new breeding lines. The objective of this study was to elucidate the patterns of genetic diversity within and among different structural groups, and measure the extent of linkage disequilibrium (LD) of 383 globally distributed rapeseed germplasm using 8,502 single nucleotide polymorphism (SNP) markers. We divided the germplasm collection into five subpopulations (P1 to P5) according to geographic and growth habit-related patterns. All subpopulations showed moderate genetic diversity (average H = 0.22 and I = 0.34). The pairwise Fst comparison revealed a great degree of divergence (Fst > 0.24) between most of the combinations. The rutabaga type showed highest divergence with spring and winter types. Higher divergence was also found between winter and spring types. Admixture model based structure analysis, principal component and neighbor-joining tree analysis placed all subpopulations into three distinct clusters. Admixed genotype constituted 29.24% of total genotypes, while remaining 70.76% belongs to identified clusters. Overall, mean linkage disequilibrium was 0.03 and it decayed to its half maximum within < 45 kb distance for whole genome. The LD decay was slower in C genome (< 93 kb); relative to the A genome (< 21 kb) which was confirmed by availability of larger haplotype blocks in C genome than A genome. The findings regarding LD pattern and population structure will help to utilize the collection as an important resource for association mapping efforts to identify genes useful in crop improvement as well as for selection of parents for hybrid breeding.
Citation: Rahman M, Hoque A, Roy J (2022) Linkage disequilibrium and population structure in a core collection of Brassica napus (L.). PLoS ONE 17(3): e0250310. https://doi.org/10.1371/journal.pone.0250310
Editor: Tzen-Yuh Chiang, National Cheng Kung University, TAIWAN
Received: March 26, 2021; Accepted: February 14, 2022; Published: March 1, 2022
Copyright: © 2022 Rahman et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All raw sequence data and variant data are available in NCBI and ENA repository. The accession id for them are PRJNA687906 (https://www.ncbi.nlm.nih.gov/biosample/17159566) and PRJEB42419 (https://www.ebi.ac.uk/eva/?eva-study=PRJEB42419), respectively.
Funding: The study was funded by the U.S. Department of Agriculture - National Institute of Food and Agriculture (Hatch Project No. ND01581).The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Rapeseed (Brassica napus L., AACC, 2n = 4x = 38), is a recent allopolyploid of polyphyletic origin that evolved from hybridization events between two parental ancestors of B. oleracea (Mediterranean cabbage, CC, 2n = 2x = 18) and B. rapa (Asian cabbage, AA, 2n = 2x = 20) [1]. Rapeseed genotypes having < 2% erucic acid in seed and < 30 μM glucosinolates in seed meal is known as canola, which is the second largest oilseed crops produced in the world after soybean [2]. Canola oil is mostly used in frying and baking, margarine, salad dressings, and many other products. Because of its fatty acid profile and the lowest amount of saturated fat among all other oils, it is commonly consumed all over the world and is considered a very healthy oil [3]. Canola oil is also rich with alpha-linolenic acid (ALA), which is associated to a lower risk of cardiovascular disease [3]. Additionally, canola is utilized as a livestock meal and is the second largest protein meal in the world after soybean [4]. Rapeseed oil has various industrial usages. The rapeseed oil, being simple alkyl esters is the best alternative to diesel fuel. It is more energy-economic and environment friendly than diesel fuel [5]. The high erucic acid content in rapeseed oil also made it suitable for using as lubricants [6] and surfactants [7]. Rapeseed expresses three growth habits, winter, spring, and semi-winter. The spring canola is planted in the early spring and harvested in the late spring of the same growing season [8]. The winter type canola is seeded in the fall, vernalized over the winter to induce flower and harvested in the summer [8]. The semi‐winter type is needed for a shorter period of vernalization to induce flower [9].
Rutabaga (Brassica napus ssp. napobrassica L.) is a cool-weather root crop, grown as table vegetable and fodder for animals [10]. Likewise rapeseed, rutabaga was also derived from natural or spontaneous hybridization between B. rapa and B. oleracea [11]. European immigrants brought rutabaga to North America [12] from its center of origin Sweden or Finland [10, 13]. Likewise most cruciferous vegetables, rutabaga bears anti-cancer properties [14] and showed considerable variability for morphology, biotic and abiotic stress resistance, seed yield and quality [10, 15].
In the United States of America (USA), the canola production increased 13.5 folds from five years average of 1991–1995 (0.11 m tons) to five years average of 2015–2019 (1.49 m tons) [16]. At the same time, canola oil consumption has increased rapidly in last few years. Statistics shows, though canola production increased, but it is not enough to meet the demand. That’s why, every year USA imports huge amount of canola oil (2.50 m tons in 2019) from other countries [2]. In USA, canola production is restricted to north-central region and North Dakota (ND) is the leading canola growing state, where 83% of US canola is grown. The North Dakota State University (NDSU) canola-breeding program could play a vital role in canola economy by developing high yielding varieties, shortening the breeding cycle and expanding canola growing acreage.
NDSU canola breeding has already developed few varieties and handsome amount of breeding populations. However, in recent years, the low genetic diversity of the parental stock is hampering the sustainability of the program. This happened because of same sets of parents has already been crossed in different combinations. The recent origin of B. napus as a species and its very recent domestication (400 years ago), as well as selection on few phenotypes (e.g. low erucic and glucosinolate acids, seed yield) also accelerated the low diversity which threatens sustainable improvement of the crop [17]. The narrow genetic diversity might also limit the prospects for hybrid breeding where complementing genepools are needed for the optimal exploitation of heterosis [18]. Therefore, we want to expand the genetic base of NDSU stock by incorporating diversified germplasms to existing collection. To shorten the breeding cycle and maximize genetic gain, we want to use cutting-edge breeding techniques such genome wide association mapping (GWAS) and marker-assisted selection. The knowledge of population structure, genetic relatedness, and patterns of linkage disequilibrium (LD) are also prime requirements for genome-wide association study (GWAS) and genome selection directed breeding strategies [19, 20]. Therefore, it is crucial to study, preserve, and even introduce genetic diversity into rapeseed since the diversity ensures the variability for biotic and abiotic stress resistance, and various agronomical and morphological traits.
We could assess the diversity of a germplasm collection by observing the phenotypic variations or genomic variations among the individuals. Before the advent of marker technology and next generation sequencing technique (NGS), crop diversity was usually assessed based on phenotypic performance. However, phenotyping is time consuming and labor intensive. Moreover, plant growth stages and environmental factors severely affect the phenotyping, results in erroneous prediction [21]. To overcome phenotyping limitations, researchers use DNA-based molecular markers for assessing the genetic diversity. Utilization of molecular markers accelerates the pre-breeding activities, as field phenotyping and pedigree information are not required [22]. Multiple genetic diversity and population structure studies, based on molecular markers [23–27], whole genome resequencing [28], transcriptome and organellar sequencing [29] have already provided information regarding genetic diversity in various B. napus collections around the world. However, the genetic diversity of the core collection maintained by the NDSU canola-breeding program has not been revealed yet. That is why; we carried out this research to explore the genetic diversity, population structure level and relatedness among the genotypes and to investigate the linkage disequilibrium (LD) and haplotype block pattern.
Materials and methods
Plant materials
A core collection of 383 rapeseed germplasm accessions was used for this study. The core is composed of 67 advanced breeding lines developed by NDSU canola breeding program, 252 germplasm accessions collected from North Central Regional Plant Introduction Station (NCRPIS), Ames, Iowa, USA and 64 varieties collected from different countries. The breeding lines are F7 generation genotypes, obtained by crossing different parents in different combinations. Initially, we collected 500 accessions from NCRPIS and phenotyped them under field conditions. No flowering occurred in case of winter type. Among them, we choose 252 relatively homogeneous genotypes for the core collection. Finally, the core collection was composed of 155 spring, 151 winter, 60 semi-winter, and 17 rutabaga types (S1 Table). The core collection is being and will be maintained through selfing. We grouped the core collection into five subpopulations (P1 to P5) according to their type and origin. Hereafter, we referred the European winter type as subpopulation-1 (P1), Asian semi-winter type as subpopulation-2 (P2), spring type NDSU genotypes (advanced breeding lines) as subpopulation-3 (P3), spring type from different countries other than NDSU breeding lines as subpopulation-4 (P4), and rutabaga type as subpopulation-5 (P5).
Genotyping and sequencing
DNA was extracted from young leaf tissue, collected from 30 days old plants. We collected three leaf samples per genotype in tubes and flash frozen in liquid nitrogen. Each sample was composed of leaves from three different plant of same genotype. Then we lyophilized leaf tissue and ground it in tubes with stainless beads using a plate shaker. Qiagen DNeasy Kit (Qiagen, CA, USA) was used for DNA extraction (3 samples per genotype) following the manufacturer’s protocol. DNA concentration was measured using a NanoDrop 2000/2000c Spectrophotometer (Thermofisher Scientific). The sample that contains good concentration of DNA was kept and other two discarded. Then we prepared the GBS library using ApekI enzyme [30]. Finally, Sequencing of the library was done at the University of Texas Southwestern Medical Center, Dallas, Texas, USA using Illumina HiSeq 2500 sequencer.
SNP calling
SNP calling was done by TASSEL 5 GBSv2 pipeline [31] was used for SNP calling using a 120-base kmer length and minimum kmer count of ten. For alignment of the reads the rapeseed reference genome [32] (available at: ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/686/985/GCF_000686985.2_Bra_napus_v2.0/) was used. The alignment was done using Bowtie 2 (version 2.3.0) alignment tool [33]. After passing all the required steps, TASSEL 5 GBSv2 pipeline yielded 497,336 unfiltered SNPs. To obtain high quality SNPs, we filtered the raw SNPs using VCFtools [34]. Filtering criteria: minor allele frequency (MAF) ≥ 0.05, missing values (max-missing) ≤ 50%, depth (minDP) ≥ 5, min-alleles = 2 and max-alleles = 2 was maintained to have bi-allelic SNPs. This filtering yielded 53,616 SNPs. To make SNP unlinked, we thinned out SNPs present within 1,000 bp distance. The SNPs that were located outside chromosomes (i.e., position unknown), were removed. As canola is a self-pollinating crop, the SNPs that were heterozygous in more than 25% of total genotypes, were also removed using TASSEL [35]. Finally, we selected 8,502 SNP markers for this study.
Data analysis
To investigate the population structure, the core collection was differentiated into clusters using STRUCTURE v2.3.4 [36] software. For this purpose, we used admixture model with various combinations of burn-in lengths (5,000 to 100,000) and Monte Carlo Markov Chain (MCMC) lengths (5,000 to 100,000). Each combination was replicated 10 times per K (K1-K10). As we grouped the collection into five subpopulations according to their type and origin, we ran each replication considering genotype assigned to specific subpopulation as well as no subpopulation i.e. genotype unassigned to any specific subpopulation. These were done to determine the parameters needed to reach convergence. We used DeltaK approach [37] to determine the ideal number of subpopulations, which was performed by Structure Harvester [38]. We also used median (MedMedK and MaxMedK) or mean (MedMeaK and MaxMeaK) estimators of the “best” K to group the subpopulations into optimum clusters [39, 40]. Ten replicates of Q matrix were assembled using CLUMPP [41] to get individual Q matrix. Structure output was visualized using the Structure Plot v2 software [42]. Principal component analysis (PCA) was conducted by covariance standardized approach in TASSEL [35]. We constructed phylogenetic tree using MEGAX program with 1,000 bootstraps [43] using neighbor-joining (NJ) algorithm. Resulting tree was displayed using FigTree V1.4.4 [44].
We calculated analysis of molecular variance (AMOVA) to partition the genetic variance among subpopulations in Arlequin3.5. To show the divergence, we calculated average pairwise between subpopulations Fst values using Arlequin3.5 [45]. Tajima’s D value of each group was calculated using MEGAX software [43]. GenAlex v6.5 [46] was used to estimate percentage of polymorphic loci, number of effective alleles, Shannon’s information index, expected heterozygosity and unbiased expected heterozygosity of each marker and subpopulation. To visualize SNP density, we developed a distribution plot of SNP using R package CMplot (available at: https://github.com/YinLiLin/R-CMplot). The polymorphism information content (PIC) of markers was calculated using software Cervus [47]. To show relatedness among individuals, we calculated kinship (IBS) matrix using software Numericware i [48] on a 1 to 2 scale. The kinship heatmap and histogram were visualized using R package ComplexHeatmap [49]. The correlation between level of relatedness (IBS coefficients) and Shannon’s information index (I) and diversity (H) was calculated in R v3.5.2 [50].
Linkage disequilibrium (LD) pattern of whole collection and different subpopulations were analyzed using PopLDdecay [51]. The mean linked LD was calculated by dividing total r2 value with total number of corresponding loci pair. In this case, r2 > 0.2 was considered only. Same procedure was followed to calculate mean unlinked LD where r2 ≤ 0.2 was considered. Haplotype block analysis was done using PLINK [52] with a window size of 5 Mb. Confidence interval (CI) method [53] was used to identify haplotype blocks with high LD. Haplotype blocks (>19 kb), observed in one subpopulation but not in the other, were considered to be subpopulation-specific block. Haplotype blocks (>19 kb) shared by more than one subpopulation, were considered to be common to corresponding subpopulations.
Results
SNP profile
We used 8,502 SNPs, covering 19 chromosomes for this study. The marker density was one per 99.5 kb. Highest number (685 SNPs, 8.06%) markers was situated on chromosome A3 and lowest (236 SNPs, 2.78%) was on chromosome A4. In terms of density, it was highest on chromosome A7 (71.1 kb) and was lowest on chromosome C9 (134.5 kb) (Table 1, Fig 1).
Frequency of SNPs varies according to color gradient.
The transition SNPs (4,956 SNPs) was more frequent than transversions (3,546 SNPs) with a ratio of 1.40. The ratio of transitions to transversions SNPs was higher in A genome (1.41) than that of in C genome (1.38). In both genome, G/C transversions were lowest (4.33% and 4.29%), but A/G and C/T transitions occurred in almost similar frequencies (Table 2). The inbreeding coefficient within individuals (Fit), inbreeding coefficient within subpopulations (Fis), observed heterozygosity (Ho) and fixation index (F) of all the markers ranged from -0.45 to 1.00, 0 to 0.73, 0 to 0.57 and 0.40 to 1.00, respectively. The mean Shannon’s information index (I) of all markers 0.37 with a range from 0.10 to 0.69. The expected heterozygosity (He) was from 0.05 to 0.50 with a mean value of 0.27. The polymorphic information content (PIC) of all markers was less than 0.50 with a mean value of 0.22 (range: 0.05 to 0.37) (S2 Table). Subpopulation-wise marker diversity parameters are presented in S3 Table.
Population structure
We did structure analysis seven times with accessions unassigned and seven times with accession assigned to their type and countries of origin. Delta K approach indicated 3 to 9 clusters (Fig 2A and 2B), while four alternative statistics (MedMedK, MedMeaK, MaxMedK, and MaxMeaK) determined following Puechmaille [39] and Li and Liu [40] indicated 3 clusters (Table 3, Fig 2C). For each run, Delta K approach showed differences in cluster number for both conditions: genotypes unassigned or assigned to their respective type and countries of origin. However, opposite scenario was found for MedMedK, MedMedK, MedMedK, and MaxMeaK statistic i.e., for each run it indicated three clusters. These outputs confirmed that Puechmaille [39] and Li and Liu [40] method was more consistent than Evanno [37] method (Table 3). Structure analysis revealed that 70.76% of genotypes belong to any of the three clusters at similarity coefficient of 0.7 and 29.24% of genotypes are admixed (Table 4, Fig 2D). Spring type accessions fall under cluster-1, whereas winter type European accessions fall under cluster-3. Cluster-2 consists of all rutabaga types and different type rapeseed accessions (Table 4). We performed principal component analysis (PCA) to show the genetic similarity among subpopulations and genotypes. The first two axes explained 21% (PCA1 13.5% and PCA2 7.22%) of the total observed variation (S4 Table). The PCA revealed that rutabaga (P5) and other types having Asian origin make one group, whereas spring type (P3, P4) and European winter type (P1) make two distinct groups (Fig 3). In addition to that, we also constructed unrooted phylogenetic tree based on neighbor joining (NJ) criteria (Fig 4). The output of neighbor-joining (NJ) tree analysis was in line with that of PCA.
2.3.4. Graphical representation of optimal number of clusters (K) determined by Evanno’s method [37] with genotypes unassigned (A) and assigned (B) to their respective countries, as well as by Puechmaille [39] and Li and Liu [40] method (C). Estimated population structure of 383 rapeseed genotypes on K = 3 (D) using Puechmaille [39] and Li and Liu [40] method.
Colors represent subpopulations.
Each branch is color-coded according to genotype belongs to subpopulation P1 to P5. Genotypes were grouped into three clusters by dividing the tree using black solid lines according to structure output.
Population diversity
Polymorphic loci percentage was greater than 75% in all subpopulations. P1 bears highest (99%) polymorphic loci, whereas it was lowest in P5 (75%). The diversity (H) was lowest in P4 and P5 (0.19) and was highest in P2 (0.25) with an average of 0.22. The Shannon’s information index (I) ranged from 0.31 (P4 and P5) to 0.40 (P2) with an average of 0.34. The Tajima’s D value ranged from -0.70 (P4) to 0.53 (P1) with an average of 0.13 (Table 5).
Population genetic differentiation
The analysis of molecular variance (AMOVA) revealed that variance among subpopulations covered 24% of total variation and rest of its was covered by among individual variance (Table 6) with a Fst and Nm value of 0.24 and 1.28, respectively.
We found significant (p < 0.01) between subpopulation Fst in all combinations. Except combinations P3 and P4 (0.11), P1 and P2 (0.19), we found Fst > 0.20 for all combinations. The pairwise Fst > 0.30 was observed between P1 and P5, P3 and P5, P4 and P5 (Table 7).
Kinship analysis showed that the IBS coefficients of the collection ranged from 1.21 to 1.94 with an average coancestry 1.47 between any two canola genotypes (Fig 5, S5 Table). Under P2 subpopulation, almost 50% of total genotypic pairs shows IBS coefficients less than 1.50. In case of other subpopulation, portion of genotypic pairs having IBS coefficient less than 1.50, was very low (Table 8, S1 Fig).
We also performed correlation analysis between mean pairwise relatedness (IBS coefficients) among individuals within subpopulation and Shannon’s information index (I), diversity (H). The I and H were significantly and negatively correlated with relatedness (r = -0.97, -0.98, and p < 0.01), respectively.
Linkage disequilibrium pattern
Subpopulation, genome, and chromosome-wise linkage disequilibrium (LD) pattern was investigated. LD = r2 values showed inverse relationship with distance i.e., mean LD was high (r2 > 0.22) at short distance bin (0–2 kb) and decreases with bin distance increment (S6 Table). In the entire collection considering both A and C genome, the mean linked LD and mean unlinked LD was 0.44 and 0.02 respectively; and loci pair under linked LD and unlinked LD was 1.81% and 98.20%, respectively. Subpopulation-wise mean linked LD ranged from r 2 = 0.41 (P2) to r 2 = 0.48 (P1). Subpopulation P5 harbored highest (8.76%) loci pair in linked LD and it was lowest in P1 (1.52%). The mean linked LD, mean LD and loci pair under linked LD was always higher in all cases in case of C genome than that of A genome (Table 9). We also compared the LD decay rate based on distance at which LD decayed to its half maximum (half-life), which is the point at which the observed r2 between sites decays to less than half the maximum r2 value. In the whole collection, LD decayed to its half maximum within < 45 kb distance for whole genome, < 21 kb for A genome, and < 93 kb for C genome. In all subpopulations, the distance for LD decay to its half maximum was always higher for C genome than A genome. LD decay rate also varied according to chromosome (S2 and S3 Figs). LD decay was lowest in chromosome C1 (348 kb) and C2 (244 kb), but was highest in chromosome A5 (13 kb) and A1 (16 kb) (S7 Table). LD decayed to its half–maximum within < 29 kb for P1, <45 kb for P2 & P3, <101 kb for P4, and <120 kb for P5. In all subpopulations, LD persisted also longest in all chromosomes of C genome than that of A genome (Fig 6, S7 Table).
We also performed haplotype block (HBs) analysis to investigate LD variation patterns across whole genome. A total 200 blocks covering 18 Mb out of the 976 Mb anchored B. napus reference genome [32], were identified. A and C genome contained 67 and 133 haplotype blocks, respectively. The total length of A and C genome specific HBs were 1.8 Mb and 16 Mb, respectively. The total length of HBs varied greatly from chromosome to chromosome. Total HBs length varies from 24 kb on A1 to 901 kb on A9 in A genome and in C genome it varies between 40 kb on C9 to 3,610 kb on C2. The haplotype block (HBs) number and size in C genome chromosome was always higher than that of A genome chromosome (Table 10). We analyzed subpopulation specific and common HBs. We found C genome chromosome bears more subpopulation specific HBs than A genome chromosome (Table 11). We also found some HBs were shared by different subpopulations, but we did not find any HBs blocks that was shared by all five subpopulations (Table 12). The shared HBs were usually located on C genome chromosome. Rutabaga type shared different HBs with other types also.
Discussion
Genotyping-by-sequencing [30] is one approach to obtain high frequency SNPs. The strategy has been used for population genetic studies, association mapping, and proven to be a powerful tool to dissect multiple genes/QTL in many plant species [54–56]. We obtained 497,336 unfiltered SNPs markers of which 8,502 high quality SNP markers were used for genetic diversity analysis of 383 genotypes. Delourme et al. (2013) [23] conducted genetic diversity analysis in B. napus using 7,367 SNP markers of 374 genotypes. However, different marker technologies such as Single Sequence Repeat (SSR), Sequence Related Amplified Polymorphism (SRAP) markers have been used by other researchers for genetic diversity analysis in B. napus. Chen et al. (2020) [57] used 30 SSR markers, Wu et al. (2014) [58] utilized 45 SSR markers, Ahmad et al. (2014) [59] used 20 SRAP markers for genetic diversity and population structure analysis of B. napus. Earlier, our group conducted a genetic diversity study of flax using 373 germplasm accessions with 6200 SNP markers [60].
The SNP markers were distributed throughout 19 chromosomes of B. napus and the marker density was one per 99.5 kb. This is comparable density to earlier study conducted by Delourme et al. (2013) [23]. Therefore, this marker density provides a sufficient resolution to estimate genome-wide diversity as well as the extent of LD within the genome. This marker density will also help in association mapping studies to identify a causal locus/loci or linked loci that can be further used either in MAS or to pinpoint the causative locus [61] especially for oligogenic traits. However, for polygenic traits such as seed yield, it is better to incorporate more markers for genome wide association studies. The core collection utilized in this study represents mostly adapted lines from various breeding programs. Therefore, sources of variation, markers of interest identified in the collection can be directly used in breeding programs.
We have identified higher frequency of transition SNPs over transversion SNPs that is an agreement with Bus et al. (2012) [62], Clarke et al. (2013) [63], and Huang et al. (2013) [64] in B. napus. Higher number of transition SNPs over transversion is also reported in other crop species such as Hevea brasiliensis [65], Camellia sinensis [66], Camelina sativa [67], and Linum usitatissimum [60].
To assess the suitability of marker for linkage analysis and diversity, we calculated PIC and expected heterozygosity (He) of markers [68]. In our research, the PIC value is ranged from 0.05 to 0.35 indicating that the markers are modestly informative. The similar lower PIC value (0.1 to 0.35) was reported by Delourme et al. (2013) [23] in B. napus. The lower PIC value is a result of bi-allelic nature of SNP markers and probable low mutation rate [69]. In our study, the He value of each marker was always greater than corresponding PIC value indicating an average lower allele frequency in the population [68].
Population diversity and structure
We have identified a moderate diversity (average H = 0.22) within the subpopulations. B. napus is capable of self-pollination, and little cross-pollination may be occurred by insect. Being a mostly self-pollinated crop a low to moderate subpopulation diversity in B. napus is expected. Low to moderate diversity was also found in previous studies [70–72]. Along with the reproduction system, one needs to look at evolution and domestication history for explaining low to moderate levels of diversity in B. napus. This allopolyploid species originated at Mediterranean coast because of a natural cross between B. rapa and B. oleracea which occurred approximately 0.12–1.37 million years ago [73, 74]. The domestication of B. napus occurred very recently, around 400 years ago with the first rapeseed being most likely a semi-winter type due to the mild climate in the region [75, 76]. Later on, European growers developed the winter and spring type Brassicas through selection for cold hardiness or early flowering to expand its cultivation in further North in the last century [77]. Therefore, the low to moderate diversity in winter and spring B. napus can be mostly explained by a recent history of the species, followed by infrequent exchange of genetic material with other Brassicas [23], as well as by the traditional breeding practices selecting for only few phenotypes. In our study, the more diversity in semi-winter type (P2, H = 0.25) than winter (P1, H = 0.21) and spring (P4, H = 0.19) type is supported by its domestication history. The Nm value was greater than one, which indicates that there was enough gene flow among semi-winter, winter, and spring types. These findings also support the evolution of winter and spring types from semi-winter type. In this research, Tajima’s D value was calculated to identify the extent of availability rare and unique alleles [78]. Recently, the NDSU canola-breeding program developed the P4 advanced breeding lines through crossing different genetic resources including winter, spring, and semi winter types and subsequent selection. This current expansion of P4 was supported by its negative Tajima’s D value, which harbors more rare alleles [79]. The subpopulation P1, P2, P3, and P5 showed positive Tajima’s D value indicating an excess of intermediate frequency alleles, which may be caused by balancing selection, population bottleneck, or population subdivision. Previously, negative Tajima’s D values were found in spring and winter type B. napus accessions [80]. The negative correlation between diversity indices (H and I) and relatedness (average IBS coefficients) indicates that subpopulation differentiation was also due to selfing and genetic drift. Flax [60] and Arapaima gigas species [81] also showed same scenario.
To exploit diversity and transgressive segregation, parents from divergent group should be crossed. Pairwise Fst statistic, a parameter describing population structure differentiation [82], was estimated among five subpopulations. In the present study all pairwise Fst values comprising both low and high values, were statistically significant. Similar results were also found in other studies [80, 83–85]. Lower pairwise Fst (0.11) was identified between spring type originated in USA (P4) and spring type originated in other countries (P3). This is reasonably justified as both subpopulations comprise of spring type genotypes and germplasm exchanged occurred between USA and other countries. It also indicating that we will not get higher genetic diversity in population if we use only spring types in the crossing program. But this combination is good for accumulating specific elite trait if the targeted trait is found in members of one and missing from the members of another group. We found spring type (P3 and P4) genotypes are greatly divergent (Fst > 0.20) from winter and semi-winter type (P1 and P2) genotypes. Utilization of genotypes from these groups in crossing program will broaden the genetic base of developed population results in high heterosis. This potentiality has already been proved as hybrids between the Chinese semi-winter and European (including Canada) spring type exhibited high heterosis for seed yield [86]. The P5 (rutabaga type) showed the higher Fst with other subpopulations such as the highest Fst was observed between P5 and P4 (NDSU spring type) followed by P3 (other spring type), P1 (winter type) and P2 (semi-winter type). This outcome clearly shows that rutabaga is genetically distinct from spring and winter type canola that is confirmed by previous studies [75, 87, 88]. This distinctness of rutabaga can be exploited through heterosis breeding. Several previous studies have already showed rutabaga as a potential gene pool for the improvement of spring canola [89, 90]. NDSU canola breeding program also utilized winter and rutabaga types in the breeding program for increasing genetic diversity and for improvement of spring canola. AMOVA showed that variation among individual within subpopulation captured greater portion of total variation, than that by among subpopulation. This finding is also supported by earlier researches [56, 72, 91, 92]. This finding supports that within subpopulation genotype from P2, P3, and P1 could be crossed as they showed high diversity (H > 0.20) for cultivar development.
Principal component analysis and distance-based population structure analysis such as NJ tree yielded three subgroups in the core collection. Here, we ran structure analysis many times to obtain convergence before the best number of clusters was determined. It was done because previous studies [39, 93] reported that STRUCTURE program did not depict the main clusters within a collection. Based on Evanno’s ΔK method [37] and MedMedK, MedMeaK, MaxMedK and MaxMeaK statistics [39], structure analysis divided the core collection into three distinct clusters. Cluster-1 contains spring type NDSU advanced breeding lines (P4) and spring type (P3) other than those. This finding is supported by low genetic differentiation (Fst = 0.11) between P3 and P4 due to sharing of parents by advanced breeding lines from P3. Cluster-3 is solely dominated by European winter type (P1) genotypes. This is also supported by high Fst between winter and other types which may be due to geographic barriers between Europe and America, Asia. Cluster-2 contained all rutabaga types as well as other type Asian genotypes which indicates that all types share considerable amount of SNP markers attributing to this cluster. These findings also indicate that there is gene flow among different types, which is also supported by Nm > 1. Structure analysis revealed that all clusters contained both non-admixed as well as admixed (share alleles attributed to different subpopulation) genotypes. For broadening genetic diversity of population, non-admixed genotypes should be crossed. However, for improving or introgression of specific traits, admixed genotypes could also be crossed which will reduce the population size required for phenotypic screening. However, structure analysis may overestimate the differentiation among individuals, as the individuals may not share alleles from same ancestors [94]. Since a breeder would like to combine historically never combined favorable alleles, IBS values directs which individuals should be crossed. Low IBS is the best. However, self-pollinated crops exhibit higher kinship values than cross-pollinated crops, as homozygosity increases probability of being identical by state [95]. We found approximately 64% of pairwise coancestry ranged from 1.21 to 1.50. Crossing among genotypes from subpopulation P2 will demonstrate more diversity, than that of other subpopulations, as most genotypic combinations of P2 shows low IBS coefficients than others do. This finding is in line with the evolutionary history of B. napus where semi-winter type is the base population containing more divergence. Gradually this diversity is narrowed down in P3 (spring type, mixed origin) and P1 (winter type), because genotypic pairs belong to P3 and P1 having high IBS values evolved from semi-winter type [77]. Subpopulation P4 exhibited highest number of pairs having IBS > 1.5, which is obvious as these genotypes are advanced breeding lines developed from crossing of same set of parents in different combinations. Genotypic pairs of P5 (rutabaga type) also showed high coancestry may be due to the duplicates which is supported by low genetic differentiation of Nordic rutabaga accessions [27]. We could discard the duplicates during the crossing program.
Linkage disequilibrium
Linkage disequilibrium can be defined as the correlation among polymorphisms in a given population [96]. The strength of association mapping relies on the degree of LD between the genotyped marker and the functional variant. Linkage disequilibrium analysis provides insight into the history of both natural and artificial selection (breeding) and can give valuable guidance to breeders seeking to diversify crop gene pools [17]. SNPs in strong LD are organized into haplotype blocks, which can extend even up to few Mb based on the species and the population used. Genetic variation across the genome is defined by these haplotype blocks. Haplotypes, which are subpopulation-specific, are defined by various demographic parameters like population structure, domestication, and selection in combination with mutation and recombination events. Conserved haplotype structure can then be used for the identification and characterization of functionally important genomic regions during evolution and/or selection [97]. In addition, the extent of LD needs to be quantified across the genome at high resolution (down to approximately one Kbp) [98]. The information is important for choosing crossing schemes, association studies and germplasm preservation strategies [99–102].
We used markers from across the genome to quantify the LD for the core collection. Low level of LD was evident for each individual subpopulation in A, C, and whole gnome. The low level of LD can be due to multiple factors. First, canola is a partially outcrossing species with an average of 21–30% of cross-pollination [103–105]. The outcrossing occurring in canola leads to more recombination and to a breakdown of haplotype blocks. Secondly, the ancestral history of canola is limited in comparison with other crops, such as rice, common bean, wheat, and corn, restricting the selection of desirable haplotypes during the evolution. In other words, there was no adaptation or domestication pressure on the species, which would lead towards positive selection. Third, the only selection pressure imposed on the species for a relatively short time was breeding. However, the breeding practices were biased towards selection of only few phenotypes. Additionally, the short period under selection pressure might have not been sufficient to select favorable haplotypes in the genome. Fourth, since canola cultivars with different growth habits are compatible there has been always gene flow present between them contributing to the low level of LD. The Nm >1 was observed in this study, which supports this gene flow. Fifth, the restriction enzyme used to develop the libraries for sequencing of the core collection helped in identification of SNPs largely residing in genic regions, which are prone to high recombination, contributing to the low level of LD. Finally, the low level of LD may be due to thinning of markers, as we did not use all markers (53,616) for LD analysis rather used 8,502 markers after thinning. That can be confirmed by analysis the LD using whole marker set in further analysis.
In this study, we have identified that the LD decay in B. napus varied across chromosomes of both A and C genomes. In addition, LD in C genome decayed much slower than A genome. C genome also contained larger haplotype blocks than A genome. This LD patterns are consistent with previous findings [17, 26, 106–108]. The slower LD decay and presence of long haplotype blocks in C genome indicates that high level of gene conservation could have resulted from limited natural recombination or could be exchanged of large chromosomal segment during evolution. In the whole genome, presence of subpopulation specific haplotype blocks suggests that these regions had been experienced selection pressure for specific geographic regions adaptation. In all subpopulations, presence of shorter haplotype blocks in A genome than C genome reveals that B. rapa progenitor of B. napus containing A genome, which has been used as oilseed crop and probably being used in hybridization process. Sharing haplotype blocks by different subpopulations especially in C genome also confirms its conserved nature. The low level of LD or haplotype blocks has implications for association mapping and a proper experimentation design is necessary for utilizing a reduced set of markers by tagging major haplotypes [109]. Though low LD of A genome requires more markers to pinpoint the location of various QTL, but once a marker is found to be significantly associated with a phenotype, there might be a higher probability of identifying the casual gene than that of C genome.
Conclusions
This study provides a new insight to select the best parents in crossing plan to maximize genetic gain in the population. The population structure analysis showed a clear geographic and growth habit related clustering. The rutabaga type showed the highest genetic divergence with spring and winter types accessions. Therefore, the breeding strategies to increase the genetic diversity may include generating population from rutabaga and spring crosses, or using rutabaga and winter crosses. The linkage disequilibrium analysis revealed the decay pattern and haplotype blocks in A and C genome. This output will help the breeder to formulate breeding strategies to develop improved cultivars using modern breeding tools by utilizing this collection and SNP markers.
Supporting information
S1 Table. List of the genotypes analyzed in this study.
https://doi.org/10.1371/journal.pone.0250310.s001
(XLSX)
S3 Table. Subpopulation-wise marker diversity parameters.
https://doi.org/10.1371/journal.pone.0250310.s003
(XLSX)
S4 Table. (a) Percentage of variation explained by the first 3 axes, (b) Eigen values by axis and sample eigen vectors.
https://doi.org/10.1371/journal.pone.0250310.s004
(XLSX)
S6 Table. Mean LD values according to distance.
https://doi.org/10.1371/journal.pone.0250310.s006
(XLSX)
S7 Table. Subpopulation-wise and chromosome-wise LD decay rate (Kb) within each subpopulation.
https://doi.org/10.1371/journal.pone.0250310.s007
(XLSX)
S2 Fig. Chromosome-wise LD decay rate (Kb) in A genome considering whole collection.
https://doi.org/10.1371/journal.pone.0250310.s009
(TIFF)
S3 Fig. Chromosome-wise LD decay rate (Kb) in C genome considering whole collection.
https://doi.org/10.1371/journal.pone.0250310.s010
(TIFF)
References
- 1. Nagaharu U. Genome-analysis in Brassica with special reference to the experimental formation of B. napus and peculiar mode of fertilization. Japanese J Bot. 1935; 7: 389–452.
- 2.
USDA Foreign Agricultural Service. Oilseeds: World markets and trade reports. 2020; https://apps.fas.usda.gov/psdonline/circulars/oilseeds.pdf. [Accessed February 20, 2020].
- 3. Connor WE. Importance of n-3 fatty acids in health and disease. American Journal of Clinical Nutrition. 2000; 1: 171S–175S. pmid:10617967
- 4. Swanepoel N, Robinson PH, Erasmus LJ. Effects of ruminally protected methionine and/or phenylalanine on performance of high producing Holstein cows fed rations with very high levels of canola meal. Anim Feed Sci Technol. 2015; 205:10–22.
- 5. Piazza GJ, Foglia TA. Rapeseed oil for oleochemical usage. Eur J lipid Sci Technol. 2001; 103(7):450–4.
- 6. Erickson DB, Bassin P. Rapeseed and cramble: alternative crops with potential industrial uses. 1990; Bull Kans Agric Exp Stn 656:1–33.
- 7. Sources Leonard C. and commercial applications of high erucic vegetable oils. Lipid Tech. 1994; 4:79–83.
- 8. Rahman M, Mcclean P. Genetic analysis on flowering time and root system in brassica napus L. Crop Sci. 2013; 53: 141–147.
- 9. Wang N, Qian W, Suppanz I, Wei L, Mao B, Long Y, et al. Flowering time variation in oilseed rape (Brassica napus L.) is associated with allelic variation in the FRIGIDA homologue BnaA.FRI.a. J Exp Bot. 2011; 62: 5641–5658. pmid:21862478
- 10.
Gowers S. Swedes and turnips. In: Bradshaw JE, editor. Root and tuber crops. Handbook of plant breeding, vol. 7. New York: Springer. 2010. p. 245–89.
- 11.
Iñiguez Luy FL, Federico ML. The genetics of Brassica napus L. In: Bancroft I, Schmidt R, editors. Genetics and genomics of the Brassicaceae. New York Dordrecht, Heidelberg, London: Springer. 2011. p. 291–322.
- 12.
Sturtevant EL. Sturtevant’s notes on edible plants. Geneva: New York Agr. Exper. Sta; 1919. p. 304–5. https://doi.org/10.1073/pnas.5.5.168 pmid:16576369
- 13.
Ahokas H. On the evolution, spread and names of rutabaga. In: MTT-agrifood research Finland 2004. Helsinki: Kave. 2004. p. 32.
- 14. Pasko P, Bukowska-Strakova K, Gdula-Argasinska J, Tyszka-Czochara M. Rutabaga (Brassica napus L. var. napobrassica) seeds, roots, and sprouts: a novel kind of food with antioxidant properties and proapoptotic potential in Hep G2 hepatoma cell line. J Med Food. 2013; 16(8):749–59. pmid:23957358
- 15. Gemmell DJ, Griffiths DW, Bradshaw JE. Effect of cultivar and harvest date on dry-matter content, hardness and sugar content of swedes for stockfeeding. J Sci Food Agric. 1990; 53(3):333–42. https://doi.org/10.1002/jsfa.2740530306.
- 16.
NASS. National Agricultural Statistics Service. 2020; https://www.nass.usda.gov. [Accessed February 20, 2020].
- 17. Qian L, Qian W, Snowdon RJ. Sub-genomic selection patterns as a signature of breeding in the allopolyploid Brassica napus genome. BMC Genomics. 2014. pmid:25539568
- 18. Girke A, Schierholt A, Becker HC. Extending the rapeseed genepool with resynthesized Brassica napus L. I: Genetic diversity. Genet Resour Crop Evol. 2012; 59: 1441–1447.
- 19. Iqbal MJ, Mamidi S, Ahsan R, Kianian SF, Coyne CJ, Hamama AA, et al. Population structure and linkage disequilibrium in Lupinus albus L. germplasm and its implication for association mapping. Theor Appl Genet. 2012; 125: 517–530. pmid:22454146
- 20. Gurung S, Mamidi S, Bonman JM, Xiong M, Brown-Guedira G, Adhikari TB. Genome-wide association study reveals novel quantitative trait loci associated with resistance to multiple leaf spot diseases of spring wheat. PLoS One. 2014; 9: e108179. pmid:25268502
- 21. Van Beuningen LT, Busch RH. Genetic diversity among North American spring wheat cultivars: III. Cluster analysis based on quantitative morphological traits. Crop Sci. 1997;37(3):981–8.
- 22. Bohn M, Utz HF, Melchinger AE. Genetic similarities among winter wheat cultivars determined on the basis of RFLPs, AFLPs, and SSRs and their use for predicting progeny variance. Crop Sci. 1999;39(1):228–37.
- 23. Delourme R, Falentin C, Fomeju BF, Boillot M, Lassalle G, André I, et al. High-density SNP-based genetic map development and linkage disequilibrium assessment in Brassica napus L. BMC Genomics. 2013; 14:120. pmid:23432809
- 24. Li F, Chen B, Xu K, Wu J, Song W, Bancroft I, et al. Genome-wide association study dissects the genetic architecture of seed weight and seed quality in rapeseed (Brassica napus L.). DNA Res. 2014; 21: 355–367. pmid:24510440
- 25. Raman H, Dalton-Morgan J, Diffey S, Raman R, Alamery S, Edwards D, et al. SNP markers-based map construction and genome-wide linkage analysis in Brassica napus. Plant Biotechnol J. 2014; 12: 851–860. pmid:24698362
- 26. Wang N, Li F, Chen B, Xu K, Yan G, Qiao J, et al. Genome‑wide investigation of genetic changes during modern breeding of Brassica napus. Theor Appl Genet. 2014; 127: 1817–1829. pmid:24947439
- 27. Yu Z, Fredua-Agyeman R, Hwang S-F, Strelkov SE. Molecular genetic diversity and population structure analyses of rutabaga accessions from Nordic countries as revealed by single nucleotide polymorphism markers. BMC Genomics. 2021; 22(1):1–13. pmid:33388042
- 28. Lu K, Wei L, Li X, Wang Y, Wu J, Liu M, et al. Whole-genome resequencing reveals Brassica napus origin and genetic loci involved in its improvement. Nat Commun. 2019; 10(1):1–12. pmid:30602773
- 29. An H, Qi X, Gaynor ML, Hao Y, Gebken SC, Mabry ME, et al. Transcriptome and organellar sequencing highlights the complex origin and diversification of allotetraploid Brassica napus. Nat Commun. 2019; 10(1):1–12. pmid:30602773
- 30. Elshire RJ, Glaubitz JC, Sun Q, Poland JA, Kawamoto K, Buckler ES, et al. A robust, simple genotyping-by-sequencing (GBS) approach for high diversity species. PLoS One. 2011; 6(5):e19379. pmid:21573248
- 31. Glaubitz JC, Casstevens TM, Lu F, Harriman J, Elshire RJ, Sun Q, et al. TASSEL-GBS: a high capacity genotyping by sequencing analysis pipeline. PLoS One. 2014; 9(2): e90346. pmid:24587335
- 32. Sun F, Fan G, Hu Q, Zhou Y, Guan M, Tong C, et al. The high-quality genome of Brassica napus cultivar ‘ZS 11’reveals the introgression history in semi-winter morphotype. Plant J. 2017; 92(3):452–68. pmid:28849613
- 33. Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012; 9(4): 357–359. pmid:22388286
- 34. Danecek P, Auton A, Abecasis G, Albers CA, Banks E, DePristo MA, et al. The variant call format and VCFtools. Bioinformatics. 2011; 27(15):2156–8. pmid:21653522
- 35. Bradbury PJ, Zhang Z, Kroon DE, Casstevens TM, Ramdoss Y, Buckler ES. TASSEL: software for association mapping of complex traits in diverse samples. Bioinformatics. 2007; 23(19):2633–5. pmid:17586829
- 36. Pritchard JK, Stephens M, Donnelly P. Inference of population structure using multilocus genotype data. Genetics. 2000; 155(2): 945–59. pmid:10835412
- 37. Evanno G, Regnaut S, Goudet J. Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Mol Ecol. 2005; 14(8):2611–20. pmid:15969739
- 38. Earl DA, vonHoldt BM. STRUCTURE HARVESTER: A website and program for visualizing STRUCTURE output and implementing the Evanno method. Conserv Genet Resour. 2012; 4: 359–361.
- 39. Puechmaille SJ. The program structure does not reliably recover the correct population structure when sampling is uneven: subsampling and new estimators alleviate the problem. Mol Ecol Resour. 2016; 16(3):608–27. pmid:26856252
- 40. Li Y-L, Liu J-X. StructureSelector: A web-based software to select and visualize the optimal number of clusters using multiple methods. Mol Ecol Resour. 2018; 18(1):176–7. pmid:28921901
- 41. Jakobsson M, Rosenberg NA. CLUMPP: a cluster matching and permutation program for dealing with label switching and multimodality in analysis of population structure. Bioinformatics. 2007; 23(14):1801–6. pmid:17485429
- 42. Ramasamy RK, Ramasamy S, Bindroo BB, Naik VG. STRUCTURE PLOT: a program for drawing elegant STRUCTURE bar plots in user friendly interface. Springerplus. 2014; 3(1):431. pmid:25152854
- 43. Kumar S, Stecher G, Li M, Knyaz C, Tamura K. MEGA X: molecular evolutionary genetics analysis across computing platforms. Mol Biol Evol. 2018; 35(6):1547–9. pmid:29722887
- 44.
Rambaut A, FigTree V. 1.4. 4. [Internet]. 2018; [Available from: http://tree.bio.ed.ac.uk/software/figtree]
- 45. Excoffier L, Lischer HEL. Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Mol Ecol Resour. 2010; 10(3): 564–7. pmid:21565059
- 46. Peakall R, Smouse PE. GenAlEx 6.5: genetic analysis in Excel. Population genetic software for teaching and research—an update. Bioinformatics [Internet]. 2012 Jul 20; 28(19):2537–9. Available from: pmid:22820204
- 47. Slate J, Marshall T, Pemberton J. A retrospective assessment of the accuracy of the paternity inference program CERVUS. Mol Ecol. 2000; 9(6):801–8. pmid:10849296
- 48. Kim B, Beavis WD. Numericware i: Identical by State Matrix Calculator. Evol Bioinforma. 2017; 13:1176934316688663. pmid:28469375
- 49. Gu Z, Eils R, Schlesner M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics. 2016; 32(18): 2847–9. pmid:27207943
- 50.
R Core Team. R: A Language and Environment for Statistical Computing [Internet]. Vienna, Austria; 2019. Available from: https://www.r-project.org/
- 51. Zhang C, Dong SS, Xu JY, He WM, Yang TL. PopLDdecay: A fast and effective tool for linkage disequilibrium decay analysis based on variant call format files. Bioinformatics. 2019; 35: 1786–1788. pmid:30321304
- 52. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MAR, Bender D, et al. PLINK: A tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007; 81: 559–575. pmid:17701901
- 53. Gabriel SB, Schaffner SF, Nguyen H, Moore JM, Roy J, Blumenstiel B, et al. The structure of haplotype blocks in the human genome. Science. 2002; 296: 2225–2229. pmid:12029063
- 54. Ganal MW, Durstewitz G, Polley A, Bérard A, Buckler ES, Charcosset A, et al. A large maize (Zea mays L.) SNP genotyping array: Development and germplasm genotyping, and genetic mapping to compare with the B73 reference genome. PLoS One. 2011; 6: e28334. pmid:22174790
- 55. Riedelsheimer C, Lisec J, Czedik-Eysenberg A, Sulpice R, Flis A, Grieder C, et al. Genome-wide association mapping of leaf metabolic profiles for dissecting complex traits in maize. Proc Natl Acad Sci U S A. 2012; 109: 8872–8877. pmid:22615396
- 56. Mandel JR, Nambeesan S, Bowers JE, Marek LF, Ebert D, Rieseberg LH, et al. Association mapping and the genomic consequences of selection in sunflower. PLoS Genet. 2013; 9: e1003378. pmid:23555290
- 57. Chen R, Shimono A, Aono M, Nakajima N, Ohsawa R, Yoshioka Y. Genetic diversity and population structure of feral rapeseed (Brassica napus L.) in Japan. PLoS One. 2020; 15: e0227990. pmid:31945118
- 58. Wu J, Li F, Xu K, Gao G, Chen B, Yan G, et al. Assessing and broadening genetic diversity of a rapeseed germplasm collection. Breed Sci. 2014; 64(4): 321–30. pmid:25914586
- 59. Ahmad R, Quiros CF, Rahman H, Swati ZA, others. Genetic diversity analyses of Brassica napus accessions using SRAP molecular markers. Plant Genet Resour. 2014; 12(1):14.
- 60. Hoque A, Fiedler JD, Rahman M. Genetic diversity analysis of a flax (Linum usitatissimum L.) global collection. BMC Genomics. 2020; 21:557. pmid:32795254
- 61. Rafalski A. Applications of single nucleotide polymorphisms in crop genetics. Current Opinion in Plant Biology. 2002; 5: 94–100. pmid:11856602
- 62. Bus A, Hecht J, Huettel B, Reinhardt R, Stich B. High-throughput polymorphism detection and genotyping in Brassica napus using next-generation RAD sequencing. BMC Genomics. 2012; 13(1):1–11. pmid:22726880
- 63. Clarke WE, Parkin IA, Gajardo HA, Gerhardt DJ, Higgins E, Sidebottom C, et al. Genomic DNA enrichment using sequence capture microarrays: a novel approach to discover sequence nucleotide polymorphisms (SNP) in Brassica napus L. PLoS One. 2013; 8(12):e81992. pmid:24312619
- 64. Huang S, Deng L, Guan M, Li J, Lu K, Wang H, et al. Identification of genome-wide single nucleotide polymorphisms in allopolyploid crop Brassica napus. BMC Genomics. 2013; 14(1):717. pmid:24138473
- 65. Mantello CC, Cardoso-Silva CB, da Silva CC, de Souza LM, Junior EJS, de Souza Gonçalves P, et al. De novo assembly and transcriptome analysis of the rubber tree (Hevea brasiliensis) and SNP markers development for rubber biosynthesis pathways. PLoS One. 2014; 9(7):e102665. pmid:25048025
- 66. Yang H, Wei C-L, Liu H-W, Wu J-L, Li Z-G, Zhang L, et al. Genetic divergence between Camellia sinensis and its wild relatives revealed via genome-wide SNPs from RAD sequencing. PLoS One. 2016; 11(3):e0151424. pmid:26962860
- 67. Luo Z, Brock J, Dyer JM, Kutchan T, Schachtman D, Augustin M, et al. Genetic diversity and population structure of a Camelina sativa spring panel. Front Plant Sci. 2019; 10:184. pmid:30842785
- 68. Shete S, Tiwari H, Elston RC. On estimating the heterozygosity and polymorphism information content value. Theor Popul Biol. 2000; 57(3): 265–71. pmid:10828218
- 69. Coates BS, Sumerford D V, Miller NJ, Kim KS, Sappington TW, Siegfried BD, et al. Comparative performance of single nucleotide polymorphism and microsatellite markers for population genetic analysis. J Hered. 2009; 100(5): 556–64. pmid:19525239
- 70. Yuan M, Zhou Y, Liu D. Genetic diversity among populations and breeding lines from recurrent selection in Brassica napus as revealed by RAPD markers. Plant Breed. 2004; 123: 9–12.
- 71. Li L, Chokchai W, Huang X, Huang T, Huang T, Li Q, et al. Comparison of AFLP and SSR for genetic diversity analysis of Brassica napus hybrids. J Agric Sci. 2011; 3: 101–110.
- 72. Gyawali S, Hegedus DD, Parkin IAP, Poon J, Higgins E, Horner K, et al. Genetic diversity and population structure in a world collection of Brassica napus accessions with emphasis on South Korea, Japan, and Pakistan. Crop Sci. 2013; 53, 1537–1545.
- 73. Morinaga T. Preliminary Note on Interspecific Hybridization in Brassica. Proc Imp Acad. 1929; 4: 620–622.
- 74. Cheung F, Trick M, Drou N, Lim YP, Park JY, Kwon SJ, et al. Comparative analysis between homoeologous genome segments of Brassica napus and its progenitor species reveals extensive sequence-level divergence. Plant Cell. 2009; 21: 1912–1928. pmid:19602626
- 75. Diers BW, Osborn TC. Genetic diversity of oilseed Brassica napus germplasm based on restriction fragment length polymorphisms. Theor Appl Genet. 1994; 88: 662–668. pmid:24186160
- 76. Gómez-Campo C, Prakash S. 2 Origin and domestication. Dev Plant Genet Breed. 1999; 4: 33–58.
- 77. Xiao Y, Chen L, Zou J, Tian E, Xia W, Meng J. Development of a population for substantial new type Brassica napus diversified at both A/C genomes. Theor Appl Genet. 2010; 121(6): 1141–50. pmid:20556596
- 78. Tajima F. Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics. 1989; 123(3):585–95. pmid:2513255
- 79. Suzuki Y. Statistical methods for detecting natural selection from genomic data. Genes and Genetic Systems. 2010; 85: 359–376. pmid:21415566
- 80. Gazave E, Tassone EE, Ilut DC, Wingerson M, Datema E, Witsenboer HMA, et al. Population genomic analysis reveals differential evolutionary histories and patterns of diversity across subgenomes and subpopulations of Brassica napus L. Front Plant Sci. 2016; 7:525. pmid:27148342
- 81. Torati LS, Taggart JB, Varela ES, Araripe J, Wehner S, Migaud H. Genetic diversity and structure in Arapaima gigas populations from Amazon and Araguaia-Tocantins river basins. BMC Genet. 2019; 20(1):13. pmid:30691389
- 82. Wright S. The interpretation of population structure by F-statistics with special regard to systems of mating. Evolution (N Y). 1965; 19(3):395–420.
- 83. Xiao Y, Cai D, Yang W, Ye W, Younas M, Wu J, et al. Genetic structure and linkage disequilibrium pattern of a rapeseed (Brassica napus L.) association mapping panel revealed by microsatellites. Theor Appl Genet. 2012; 125, 437–447. pmid:22437490
- 84. Liu S, Fan C, Li J, Cai G, Yang Q, Wu J, et al. A genome-wide association study reveals novel elite allelic variations in seed oil content of Brassica napus. Theor Appl Genet. 2016; 129: 1203–1215. pmid:26912143
- 85. Chen R, Hara T, Ohsawa R, Yoshioka Y. Analysis of genetic diversity of rapeseed genetic resources in Japan and core collection construction. Breed Sci. 2017; 67: 239–247. pmid:28744177
- 86. Qian W, Sass O, Meng J, Li M, Frauen M, Jung C. Heterotic patterns in rapeseed (Brassica napus L.): I. Crosses between spring and Chinese semi-winter lines. Theor Appl Genet. 2007; 115(1):27–34. pmid:17453172
- 87. Hasan M, Seyis F, Badani AG, Pons-Kühnemann J, Friedt W, Lühs W, et al. Analysis of genetic diversity in the Brassica napus L. gene pool using SSR markers. Genet Resour Crop Evol. 2006; 53: 793–802.
- 88. Bus A, Körber N, Snowdon RJ, Stich B. Patterns of molecular variation in a species-wide germplasm set of Brassica napus. Theor Appl Genet. 2011; 123: 1413–1423. pmid:21847624
- 89.
Flad DWF. Use of Rutabaga (Brassica napus var. napobrassica) for the Improvement of Canadian Spring Canola (Brassica napus). [Master’s thesis]. [Alberta (CA)]: University of Alberta. 2016; https://doi.org/10.7939/R3319S744
- 90. Shiranifar B, Hobson N, Kebede B, Yang RC, Rahman H. Potential of rutabaga (Brassica napus var. napobrassica) gene pool for use in the breeding of B. napus canola. Crop Sci. 2020; 60: 157–171.
- 91. Cruz VM V., Luhman R, Marek LF, Rife CL, Shoemaker RC, Brummer EC, et al. Characterization of flowering time and SSR marker analysis of spring and winter type Brassica napus L. germplasm. Euphytica. 2007; 153: 43–57.
- 92. Malmberg MM, Shi F, Spangenberg GC, Daetwyler HD, Cogan NOI. Diversity and genome analysis of australian and global oilseed Brassica napus L. Germplasm using transcriptomics and whole genome re-sequencing. Front Plant Sci. 2018; 9:508. pmid:29725344
- 93. Kalinowski ST. The computer program STRUCTURE does not reliably identify the main genetic clusters within species: simulations and implications for human population structure. Heredity (Edinb). 2011;106(4):625–32. pmid:20683484
- 94. Jakobsson M, Edge MD, Rosenberg NA. The relationship between FST and the frequency of the most frequent allele. Genetics. 2013; 193: 515–528. pmid:23172852
- 95. Bernardo R, Romero-Severson J, Ziegle J, Hauser J, Joe L, Hookstra G, et al. Parental contribution and coefficient of coancestry among maize inbreds: pedigree, RFLP, and SSR data. Theor Appl Genet. 2000; 100(3–4): 552–6.
- 96.
Goode EL. Linkage Disequilibrium. In: Schwab M, editor. Encyclopedia of Cancer [Internet]. Berlin, Heidelberg: Springer Berlin Heidelberg; 2011. p. 2043–8. Available from: https://doi.org/10.1007/978-3-642-16483-5_3368.
- 97. Guryev V, Smits BMG, Van De Belt J, Verheul M, Hubner N, Cuppen E. Haplotype block structure is conserved across mammals. PLoS Genet. 2006; 2: e121. pmid:16895449
- 98. Reich DE, Cargili M, Boik S, Ireland J, Sabeti PC, Richter DJ, et al. Linkage disequilibrium in the human genome. Nature. 2001; 411: 199–204. pmid:11346797
- 99. Kim S, Plagnol V, Hu TT, Toomajian C, Clark RM, Ossowski S, et al. Recombination and linkage disequilibrium in Arabidopsis thaliana. Nat Genet. 2007; 39: 1151–1155. pmid:17676040
- 100. Sachs MM. Cereal germplasm resources. Plant Physiology. 2009; 149: 148–151. pmid:19126707
- 101. Brachi B, Morris GP, Borevitz JO. Genome-wide association studies in plants: The missing heritability is in the field. Genome Biology. 2011; 12: 232. pmid:22035733
- 102. Jia G, Huang X, Zhi H, Zhao Y, Zhao Q, Li W, et al. A haplotype map of genomic variations and genome-wide association studies of agronomic traits in foxtail millet (Setaria italica). Nat Genet. 2013; 45: 957–961. pmid:23793027
- 103. Rakow G, Woods DL. Outcrossing in rape and mustard under saskatchewan prairie conditions. Can J Plant Sci. 1987; 67: 147–151.
- 104. Becker HC, Damgaard C, Karlsson B. Environmental variation for outcrossing rate in rapeseed (Brassica napus). Theor Appl Genet. 1992; 84: 303–306. pmid:24203188
- 105. Cuthbert JL, McVetty PBE. Plot-to-plot, row-to-row and plant-to-plant outcrossing studies in oilseed rape. Can J Plant Sci. 2001; 81: 657–664.
- 106. Wu Z, Wang B, Chen X, Wu J, King GJ, Xiao Y, et al. Evaluation of linkage disequilibrium pattern and association study on seed oil content in Brassica napus using ddRAD sequencing. PLoS One. 2016; 11: e0146383. pmid:26730738
- 107. Zhou Q, Zhou C, Zheng W, Mason AS, Fan S, Wu C, et al. Genome-wide SNP markers based on SLAF-seq uncover breeding traces in rapeseed (Brassica napus L.). Front Plant Sci. 2017; 8: 648. pmid:28503182
- 108. Gao H, Ye S, Wu J, Wang L, Wang R, Lei W, et al. Genome-wide association analysis of aluminum tolerance related traits in rapeseed (Brassica napus L.) during germination. Genet Resour Crop Evol. 2020; pmid:33505123
- 109. Collins FS, Green ED, Guttmacher AE, Guyer MS. A vision for the future of genomics research. Nature. 2003; 422: 835–847. pmid:12695777