Integration of Two Diploid Potato Linkage Maps with the Potato Genome Sequence

Abstract

To facilitate genome-guided breeding in potato, we developed an 8303 Single Nucleotide Polymorphism (SNP) marker array using potato genome and transcriptome resources. To validate the Infinium 8303 Potato Array, we developed linkage maps from two diploid populations (DRH and D84) and compared these maps with the assembled potato genome sequence. Both populations used the doubled monoploid reference genotype DM1-3 516 R44 as the female parent but had different heterozygous diploid male parents (RH89-039-16 and 84SD22). Over 4,400 markers were mapped (1,960 in DRH and 2,454 in D84, 787 in common) resulting in map sizes of 965 (DRH) and 792 (D84) cM, covering 87% (DRH) and 88% (D84) of genome sequence length. Of the mapped markers, 33.5% were in candidate genes selected for the array, 4.5% were markers from existing genetic maps, and 61% were selected based on distribution across the genome. Markers with distorted segregation ratios occurred in blocks in both linkage maps, accounting for 4% (DRH) and 9% (D84) of mapped markers. Markers with distorted segregation ratios were unique to each population with blocks on chromosomes 9 and 12 in DRH and 3, 4, 6 and 8 in D84. Chromosome assignment of markers based on linkage mapping differed from sequence alignment with the Potato Genome Sequencing Consortium (PGSC) pseudomolecules for 1% of the mapped markers with some disconcordant markers attributable to paralogs. In total, 126 (DRH) and 226 (D84) mapped markers were not anchored to the pseudomolecules and provide new scaffold anchoring data to improve the potato genome assembly. The high degree of concordance between the linkage maps and the pseudomolecules demonstrates both the quality of the potato genome sequence and the functionality of the Infinium 8303 Potato Array. The broad genome coverage of the Infinium 8303 Potato Array compared to other marker sets will enable numerous downstream applications.

About the Authors:

Kimberly J. Felcher

Affiliation: Department of Crop and Soil Sciences, Michigan State University, East Lansing, Michigan, United States of America

Joseph J. Coombs

Affiliation: Department of Crop and Soil Sciences, Michigan State University, East Lansing, Michigan, United States of America

Alicia N. Massa

Affiliation: Department of Plant Biology, Michigan State University, East Lansing, Michigan, United States of America

Candice N. Hansey

Affiliation: Department of Plant Biology, Michigan State University, East Lansing, Michigan, United States of America

John P. Hamilton

Affiliation: Department of Plant Biology, Michigan State University, East Lansing, Michigan, United States of America

Richard E. Veilleux

Affiliation: Department of Horticulture, Virginia Tech, Blacksburg, Virginia, United States of America

C. Robin Buell

Affiliation: Department of Plant Biology, Michigan State University, East Lansing, Michigan, United States of America

David S. Douches

* E-mail: [email protected]

Affiliation: Department of Crop and Soil Sciences, Michigan State University, East Lansing, Michigan, United States of America

Introduction

World-wide, potato is one of the most important food crops ranking third in total production behind wheat and rice [1]. However, cultivated potato (Solanum tuberosum) is an autotetraploid (2n = 4x = 48) which complicates both genetic/genomic studies as well as breeding efforts to improve important traits such as disease/pest resistance, processing quality and nutritional value. Multiple linkage maps have been constructed for potato in an effort to better understand the potato genome, develop markers for marker assisted breeding, and facilitate map-based cloning [2], [3], [4], [5]. Most potato linkage maps have been generated from diploid populations to simplify genetic segregation and to incorporate polymorphism from wild species and primitive cultivars. These maps range in size from 606 cM [2] to 1120 cM [4] and contain as few as 85 markers [6] and as many as 10,000 markers [5]. Potato linkage maps have been constructed from many types of markers, including isozymes, Restriction Fragment Length Polymorphisms (RFLPs), Simple Sequence Repeats (SSRs), Amplified Fragment Length Polymorphisms (AFLPs) and more recently, Single Nucleotide Polymorphisms (SNPs) [2], [7], [5], [8]. High frequencies of SNPs have been demonstrated in several crop species including maize [9], soybean [10] and rice [11] providing abundant variability to develop markers for marker assisted breeding.

In the last decade, the rapid evolution of next-generation sequencing technologies and associated bioinformatic pipelines has made it possible both technically and financially to generate genome sequences for many agronomically important crop species including rice (Oryza sativa) [12], maize (Zea mays) [13], soybean (Glycine max) [14] and potato (Solanum tuberosum) [15]. High density SNP maps have been developed for rice [16] and maize [17] which has increased our understanding of genome influence on crop performance. Thus, the publication of the potato genome sequence and the subsequent development and release of the Infinium 8303 Potato Array [18], [19] provide various opportunities to improve our understanding of the structure and function of the potato genome and to bridge the gap between genomics and applied breeding. We have genotyped two diploid potato populations using the Infinium 8303 Potato Array and created two linkage maps to validate the array. In addition, concordance between our linkage maps and the 12 pseudomolecules constructed by the Potato Genome Sequencing Consortium (PGSC) was tested in order to validate the potato genome sequence.

Results and Discussion

Fabrication of the Infinium 8303 Potato Array

SNPs for the Infinium array were selected from a set of biallelic, high confidence SNPs identified from transcriptome sequencing of six cultivated potato cultivars using either Sanger or Illumina transcriptome sequencing [18]. As we used SNPs derived from transcriptome not genome sequence data from only six cultivars, we can not assess allele frequency in SNP selection. Initial design of the array included 10,000 SNPs; however, due to allocation of two probes to detect transversion SNPs and assay failures, the final version of the “Infinium 8303 Potato Array” comprised 8,303 functional markers including 3,018 from candidate genes of interest, 536 from potato genetic markers and 4,749 selected for maximum genome coverage (Table S1). SNPs were distributed across the 12 chromosomes (Figure S1), providing abundant representation of the genes on all 12 chromosomes. Some SNPs (697) were on superscaffolds not anchored to the genome sequence and thus not present within the 12 pseudomolecules generated by the PGSC and used in this study. The pseudomolecules are representations of the 12 potato chromosomes generated by stitching superscaffolds, scaffolds and contigs into a contiguous sequence anchored to the chromosomes using both in silico and genetic mapping data as described previously [15]. SNPs (6,351 total) were distributed among 3,591 annotated genes, with a maximum of 29 SNPs occurring in a single gene (Table S2). Of the SNPs within annotated PGSC v3.4 genes [15], 5,538 SNPs were within the coding sequence (CDS) and 817 were within annotated untranslated regions (UTRs). The remainder of the SNPs (1,524) occurred in unannotated genes predicted from transcriptome sequencing and thus may represent genes missing within the PGSC v3.4 gene annotation dataset.

Development of two diploid segregating populations

Two segregating, diploid potato populations (F₁) derived from crosses with DM1-3 516 R44 (hereafter referred to as DM) were genotyped with the Infinium 8303 Potato Array. The female parent for both populations was DM, the doubled monoploid line used to generate the potato genome sequence [15]. DM is a homozygous diploid (2n = 2x = 24) derived from a heterozygous accession of Solanum tuberosum Group Phureja [20] using anther culture to generate a monoploid (2n = 1x = 12); leaf discs of the monoploid were subsequently placed in a tissue culture callus/regeneration protocol to induce spontaneous chromosome doubling resulting in a homozygous doubled monoploid. By using DM to generate a draft potato genome sequence, the PGSC was able to overcome the challenge of heterozygosity with respect to genome assembly. The use of DM as a parent in our diploid population also simplified the mapping process. Because DM is homozygous, all informative meioses occurred in the male parent. Therefore, a single linkage map was generated for each population representing the marker order in each male parent.

Population D84 (n = 92, from Michigan State University) resulted from a cross between DM and 84SD22. Breeding line 84SD22 (also called T704, [2], [21]) is a heterozygous diploid (S. tuberosum×S. chacoense hybrid, Figure S2) that was shown to have a higher percentage (59%) of polymorphic SNPs than other diploid clones crossed to DM in the Michigan State University potato breeding program (data not shown). Population DRH (n = 92, from Virginia Tech) was derived from a cross between DM and RH89-039-16 (hereafter referred to as RH). Breeding line RH is a heterozygous diploid with both S. tuberosum Groups Tuberosum and Phureja in its pedigree (see [15] supplemental data). The DRH population was selected for mapping because the RH clone has been used in genome sequencing [15] and because several genetic maps for RH exist [5], [8], [22], permitting comparisons to our SNP-based map of RH.

Utility of the Infinium 8303 Potato Array

In general, the number of mapped markers per chromosome corresponded to chromosome size with the largest chromosome (1) represented by the greatest number of markers (Table 1). One obvious departure from this trend was chromosome 3 in DRH with only 88 mapped markers. The proportion of mapped markers within candidate genes, previously mapped genetic markers or at random genome locations, was similar in both populations and corresponded to the distribution of markers in the Infinium 8303 Potato Array (Table 2).

[Figure omitted. See PDF.]

Table 1. Comparison of markers mapped in two diploid populations, DRH and D84.

https://doi.org/10.1371/journal.pone.0036347.t001

[Figure omitted. See PDF.]

Table 2. Distribution of mapped markers in DRH, D84 and the Infinium 8303 Potato Array.

https://doi.org/10.1371/journal.pone.0036347.t002

After filtering to remove non-informative markers (Table S3), we observed 1,989 and 2,454 segregating markers in the DRH and D84 populations, respectively. Because the female parent in both populations (DM) was homozygous, all informative meioses occurred in the male parents (RH and 84SD22) and are equivalent in number to the number of segregating markers. However, pair-wise comparisons revealed 1,031 and 1,817 co-segregating loci in the DRH and D84 populations, respectively. After the removal of co-segregating loci and assignment of markers to linkage groups, the final number of unique mapped loci for each population was 944 (DRH) and 637 (D84) (Table 1). Including the co-segregating markers, over 4,400 markers were mapped including 787 markers common to both populations (Table 1). Map sizes were 965 cM (DRH) and 792 cM (D84) (Table 1). Previous map sizes for the RH clone ranged from 773 cM [5] to 857 cM [22] both smaller than the SNP-based map of RH generated in this study (Table 3). This was likely due to greater genome coverage with the SNP array as the Infinium 8303 Potato Array was designed to cover approximately 650 Mb of potato genome sequence. Although the potato genome sequence is 727 Mb, the DM pseudomolecules are smaller (647.2 Mb) (Table 1) as not all scaffolds could be anchored to a chromosome. Thus, the DRH map (633.8 Mb) covers 98% of the pseudomolecule length and 87% of the genome sequence length. Similarly, the D84 map (641.6 Mb) covers 99% of the pseudomolecule length and 88% of the genome sequence length. Van Os et al. [5] noted a strong clustering of markers in their map of the RH clone which resulted in uneven genome coverage. Furthermore, the 944 distinct marker loci comprising our DRH SNP map represented nearly double the number of unique loci (549 recombination bins) generated by van Os et al. [5] (Table 3). Thus, using the Infinium 8303 Potato Array, we were able to generate a larger map, with more unique loci and greater genome coverage than previous maps.

[Figure omitted. See PDF.]

Table 3. Comparison of four diploid potato linkage maps: D84 and DRH maps (SNP based) and the ultra-high density (UHD) maps (AFLP-based [5]).

https://doi.org/10.1371/journal.pone.0036347.t003

The D84 linkage map was smaller than the DRH map despite the fact that it covered roughly the same proportion of the genome (Table 1, Figure 1). In a comparison of the two maps using the 787 markers that were common to both, we estimated the DRH linkage map to be 890 cM whereas the D84 linkage map was 641 cM (28% smaller than DRH) (Table S4). These data imply reduced recombination in D84. The male parent of this population was 84SD22, an S. tuberosum×S. chacoense hybrid that has a 30% reduction in recombination based on gene centromere mapping [21]. Bonierbale et al. [2] demonstrated reduced recombination in a different 84SD22-derived population. Gebhardt et al. [3] also noted significant reduction in map length with inter-specific compared to intra-specific crosses, which was attributed to reduced recombination. Therefore, the smaller linkage map we observed for the D84 population compared to the DRH population (which derived from cultigens within S. tuberosum) follows previous trends.

[Figure omitted. See PDF.]

Figure 1. Alignment of chromosome 1 from DRH (left) and D84 (right).

https://doi.org/10.1371/journal.pone.0036347.g001

Distorted segregation ratios are common in potato mapping studies with a wide range in the reported percentage of mapped loci exhibiting aberrant segregation ratios: 25% [6], 28% [2] and 27 to 40% [3]. The percentage of mapped markers with distorted segregation was 4% and 9% in the DRH and D84 populations, respectively (Table 4). However, the same calculation performed with entire marker sets (i.e. mapped markers plus co-segregating markers) resulted in an increase in the percentage of markers with distorted segregation (DRH = 6% and D84 = 21%). This difference was due to co-segregating markers at locations where distorted segregation occurred and was more pronounced in the D84 population. Despite which markers were used in the analysis, the D84 population had a greater percentage of markers with distorted segregation. A similar phenomenon was observed between an S. tuberosum-derived population (27% of loci with distorted segregation) and an inter-specific population with Solanum spegazzinii in its pedigree (40% of loci with distorted segregation) [3]. In their study, Gebhardt et al. [3] suggested that inter-specific hybrids suffer not only from reduced recombination but also preferential selection for certain allelic combinations resulting in distorted segregation. As we cannot identify the origin (S. tuberosum or S. chacoense) of the alleles in the D84 population, we cannot determine if there is preferential selection of alleles based on species of origin. Hybrid breakdown (resulting in seedling death) has been observed among S. chacoense hybrids [23] and an S. chacoense mutant causing death at the cotyledon stage was also identified [24]; both are phenomena which could contribute to distorted segregation. In our study, any seedlings that did not survive and produce minitubers were not included in the mapping population. Therefore, we cannot rule out the contribution of hybrid breakdown or mutant alleles to the distorted segregation ratios that were observed.

[Figure omitted. See PDF.]

Table 4. Markers with distorted segregation ratios in populations DRH and D84.

https://doi.org/10.1371/journal.pone.0036347.t004

The markers with distorted segregation were unique to each population and were not localized to the same genomic regions for the two populations. The majority of the markers with distorted segregation were found in blocks on chromosomes 9 and 12 in DRH and on chromosomes 3, 4, 6 and 8 in D84 (Table 4). The size of these blocks ranged from 0 cM (blocks of co-segregating loci on chromosome 4 in D84) to 19.5 cM (Table 5). Individual markers with distorted segregation were located both at the ends of and within the chromosome maps. Bonierbale et al. [2] also found blocks of markers with distorted segregation on chromosomes 6 and 8 in an S. chacoense-derived population, consistent with what was observed in the D84 population.

[Figure omitted. See PDF.]

Table 5. Position and size of blocks with distorted segregation.

https://doi.org/10.1371/journal.pone.0036347.t005

Validation of the PGSC pseudomolecules

In order to compare the maps for both populations, we graphed the DRH marker positions against D84 marker positions for the 787 markers common to both populations. If marker order was identical in both populations, we would expect a straight line with a slope of 1. This is largely the case with the exceptions of chromosomes 5 and 12 (Figure 2). For chromosome 12, there were few data points. This may have contributed to the poor correlation and obscured the cause for the lack of concordance. By inspecting the cM position of common markers on chromosome 5, we identified a block of markers in D84 that was both mis-oriented and placed at the wrong end of the chromosome compared to DRH. Bonierbale et al [2] also detected an inversion on chromosome 5 in an S. chacoense-derived map relative to the tomato map. Although we initially assumed that the D84 map was incorrect, further inspection of the data showed that the D84 map order reflected the order and orientation of the current pseudomolecules. Thus, it appears that the genome sequence at this location differs between the RH clone, which is derived from cultivated potato, and the 84SD22 clone, which is derived from the wild species S. chacoense. Hu et al. [25] also noted chromosome rearrangements between Arabidopsis thaliana and A. lyrata despite greater than 80% sequence identity between them. One should, therefore, exercise caution when extrapolating sequence data between species.

[Figure omitted. See PDF.]

Figure 2. Graphs of DRH SNP marker position (cM) vs. D84 SNP marker position (cM) for all 12 potato chromosomes.

https://doi.org/10.1371/journal.pone.0036347.g002

When marker locations on the linkage maps were compared with their assigned position on the potato pseudomolecules, some discordances were noted (Table 6 and Table S5). The chromosome assignment of 1.2% (DRH) and 1.8% (D84) of the markers differed between the linkage map and the pseudomolecule. If a discrepancy between the linkage map and the physical assemblies included more than one marker, we inspected the positions of these markers in the superscaffolds to determine whether the observed incongruity reflected a mis-assembly (i.e., incorrect order and/or orientation of superscaffolds) in the construction of the pseudomolecules. None of the disconcordant markers could be attributed to errors in the construction of the pseudomolecules. Disconcordant markers were also viewed in GenomeStudio to determine if the SNP assay was good (i.e., tight progeny clusters around the parental genotypes) or if there were multiple progeny clusters which could be indicative of a paralog. Of the 71 disconcordant markers, four (DRH) and five (D84) may be attributed to a paralog (Table S5). The presence of paralogs could also explain why some markers included in the linkage maps (7% in DRH and 10% in D84) could not be anchored to the pseudomolecules (Table 7). Among the disconcordant markers, two and 22 had distorted segregation ratios in DRH and D84, respectively, and the annotation was enriched for genes of unknown function (44%) (Table S5).

[Figure omitted. See PDF.]

Table 6. Summary of disconcordant markers for each of the mapping populations.

https://doi.org/10.1371/journal.pone.0036347.t006

[Figure omitted. See PDF.]

Table 7. Mapped markers that were not anchored to the pseudomolecules.

https://doi.org/10.1371/journal.pone.0036347.t007

By comparing the genetic location (cM) with the physical position (Mb) of each marker, we evaluated the concordance between the genetic and the physical maps and estimated genome-wide recombination rates. For most of the chromosomes, the resulting graphs had the expected shape, correlating well with chromosome structure (Figure 3). Exceptions included DRH chromosomes 3, 5 and 12 and D84 chromosome 10 which indicated local inversions and/or mis-ordering of the superscaffolds in the potato genome sequence. Further examination of the data confirmed several instances where the orientation and/or order of the superscaffolds were in error (Table S6). If data from both populations indicated an error in the superscaffolds, the order was corrected. Correcting these issues in our data sets resulted in improved graphs with the previously mentioned exception of chromosome 5 (Figure 4, Figure S3 and Figure S4). As expected, graphical representation of genome wide recombination rates (RR = cM/Mb) showed greater recombination rates near the ends of the chromosomes (Figure 3, Figure S3 and Figure S4).

[Figure omitted. See PDF.]

Figure 3. Graph of chromosome 4 (D84) showing the genetic location (cM) and the physical position (Mb) of 204 markers, and the estimated local recombination.

https://doi.org/10.1371/journal.pone.0036347.g003

[Figure omitted. See PDF.]

Figure 4. Graph of chromosome 10 (D84) showing the genetic location (cM) and the physical position (Mb) of 98 markers, and the estimated local recombination.

A. Chromosome 10 from population D84 demonstrating poor concordance between physical and genetic position. B. Chromosome 10 from population D84 after correcting the order and orientation of the superscaffolds in the pseudomolecules.

https://doi.org/10.1371/journal.pone.0036347.g004

The results presented here demonstrate the utility of the Infinium 8303 Potato Array for mapping studies, providing excellent coverage of the potato genome (633 to 641 Mb) with over 4,400 mapped markers. The high-throughput nature of the array coupled with greater genome coverage than other marker types, make this a valuable tool for Quantitative Trait Locus (QTL) analysis, Genome Wide Association Studies (GWAS) and map-based gene cloning. Furthermore, the close consensus between the two linkage maps and the genome sequence confirmed the high quality of the potato genome sequence. The identification of mis-oriented scaffolds will serve to further refine the genome sequence.

Materials and Methods

SNP Array Development

SNPs for the Infinium 8303 Potato Array were selected from the 69,011 high confidence SNPs described in Hamilton et al. [18]. From this set, 3,856 SNPs were selected based on location within a candidate gene of interest as defined by the potato community or a potato genetic marker used in previous mapping studies (Tables S1 and S2). The remaining 5,324 SNPs on the array were selected from the 69,011 high confidence SNP set using a custom Perl script to provide coverage of the genome; in total, 650 Mb of the potato genome are represented by SNPs on the array. The Infinium 8303 Potato Array was fabricated by Illumina (Illumina, San Diego, CA).

Development of a Custom Cluster File for the Infinium 8303 Potato Array

In order to test the utility of the auto-clustering function of the Illumina GenomeStudio software (Illumina, Inc., San Diego, CA), we genotyped DNA from a set of diverse potato clones (443) on an Illumina iScan Reader utilizing the Infinium® HD Assay Ultra (Illumina, Inc., San Diego, CA) and the Infinium 8303 Potato Array. Results were analyzed with the Illumina GenomeStudio software (Illumina, San Diego, CA). Due to variable cluster positioning and quality for each SNP, the software auto-clustering was not used. Clusters for the three marker classes (AA, AB, and BB) were manually positioned within GenomeStudio to generate a custom cluster file, available at http://solcap.msu.edu/potato_infinium.shtml. Note that using the broad set of germplasm when determining the cluster positions and assay quality allowed for more accurate determination of the cluster positions. In addition, the quality of each SNP was manually determined and, of the 8,303 SNPs, 637 had low signal intensity, loose clustering, or other assay failures and were removed from future analyses.

SNP Genotyping and Linkage Mapping

DNA from all parents and progeny was extracted from young leaf tissue using Qiagen DNeasy Plant Mini Kits (Qiagen, Germantown, MD), quantified using the Quant-iT™ PicoGreen® dsDNA Assay Kit (Invitrogen, San Diego, CA) and adjusted to a concentration of 50 ng·µL⁻¹. SNP genotyping with the Infinium 8303 Potato Array was performed as described above and the custom cluster file was used to generate genotype scores for the D84 and DRH mapping populations in GenomeStudio (Illumina, San Diego, CA).

Prior to mapping, SNPs were filtered to remove those which were non-informative including: SNPs that were of low quality (Table S1) or that aligned to multiple locations in the superscaffolds generated by the PGSC [15], SNPs for which parental replicated genotypes differed, SNPs for which one or both parents lacked a genotype, SNPs for which DM appeared to be heterozygous, SNPs for which both parents were homozygous (AA×AA or AA×BB), SNPs with a no-call rate of ≥12% (greater than 10 progeny with missing genotypes), SNPs for which the progeny showed no or highly distorted segregation (based on chi-square tests with Bonferroni correction for multiple comparisons) and SNPs which co-segregated with other SNPs based on pairwise comparisons. The remaining SNPs were mapped using JoinMap4 [26]. The data were entered as a cross-pollinated population type with <lmxll> segregation and markers were grouped by regression mapping using Haldane's function. Markers assigned to linkage groups had a minimum LOD score of 3 and a maximum of 10.

A custom version of the version 2.1.10 pseudomolecules was made to remove the large, Mb-sized gaps on a subset of the chromosomes (1, 2, 5 and 12) that reflected estimates of the centromeric gaps. These gaps were resized to the standard gap size of 50 kb. The 8303 SNPs were then aligned to the customized version 2.1.10 pseudomolecules with the est2genome model within exonerate version 2.2.0 [27] using 50 bp of context sequence on both sides of the SNP, a minimum intron size of 10 and a maximum intron size of 15,000. Alignments were required to have greater than 95% sequence identity, greater than 95% coverage, no insertions or deletions, and two or fewer alignments meeting these criteria per SNP. Linkage groups (LGs) generated in JoinMap4 were then aligned to the custom version of the pseudomolecules. If there was a discrepancy between the marker position on the linkage map and the corresponding SNP position on the pseudomolecules, the marker was examined in GenomeStudio to see if the graph of the genotypes indicated the possibility of a paralog.

Genome-wide recombination rates were estimated from the comparison of the linkage maps and the potato pseudomolecules. At any given nucleotide coordinate (SNP marker), the recombination rate was calculated by locally adjusting a polynomial curve to the plot of genetic versus physical distances. This was performed using the MareyMap packages, an R-based tool that uses Tcl/Tk to build the graphical interface [28]. The plots were generated with the graphical interface MareyMapGUI and the slope of the curve was obtained using the “loess” (or lowess for LOcally WEighted Scatterplot Smoothing) interpolation method.

Supporting Information

[Figure omitted. See PDF.]

Figure S1.

SNP and gene frequency distribution across potato chromosomes. Frequency is expressed in number of occurrences/100 kbp. A total number of 6,955 SNPs from the Infinium 8303 Potato Array (left) and a total number of 6,351 genes (right) were plotted against physical position to assess SNP coverage on chromosomes 1 through 12. The remaining SNPs (1,348) were not graphed because they were either not mapped to the pseudomolecules or mapped to more than one position on the pseudomolecules. If a gene did not have an associated SNP (1,952) it was not graphed (see Table S1). A. Chromosomes 1 through 6. B. Chromosomes 7 through 12.

https://doi.org/10.1371/journal.pone.0036347.s001

(TIFF)

Figure S2.

Pedigree of potato breeding line 84SD22.

https://doi.org/10.1371/journal.pone.0036347.s002

(TIFF)

Figure S3.

Graphs of the 12 DRH chromosomes showing the genetic location (cM) and the physical position (Mb) of markers, and the estimated local recombination. Physical marker position (based on corrected superscaffold ordering and orientation) was plotted against genetic marker position to identify areas of discordance between the two (as indicated by peaks and valleys in the graphs). Global recombination rates (cM/Mb) were plotted against physical position to identify areas of higher and lower recombination.

https://doi.org/10.1371/journal.pone.0036347.s003

(TIFF)

Figure S4.

Graphs of the 12 D84 chromosomes showing the genetic location (cM) and the physical position (Mb) of markers, and the estimated local recombination. Physical marker position (based on corrected superscaffold ordering and orientation) was plotted against genetic marker position to identify areas of discordance between the two (as indicated by peaks and valleys in the graphs). Global recombination rates (cM/Mb) were plotted against physical position to identify areas of higher and lower recombination.

https://doi.org/10.1371/journal.pone.0036347.s004

(TIFF)

Table S1.

Metadata for the Infinium 8303 Potato Array. SNP ID numbers with their corresponding superscaffold ID, superscaffold position, pseudomolecule ID, pseudomolecule position, gene and annotation (where applicable), SNP quality call and SNP context sequence. SNPs were manually curated to determine the quality. A SNP was discarded based on low signal intensity, loose clustering, or other assay failures. The context sequence represents the sequences submitted to Illumina. For some SNPs the reverse complement was used in the assay. UM = unmapped; a SNP could not be mapped to the superscaffolds or the pseudomolecules, MM = multiple mapping, a SNP mapped to more than one location on the superscaffolds and pseudomolecules; NG = the SNP is not in a gene.

https://doi.org/10.1371/journal.pone.0036347.s005

(XLSX)

Table S2.

List of potato genes containing SNPs on the Infinium 8303 Potato Array. PGSC gene ID numbers for genes that are represented on the Infinium 8303 Potato Array, the number of SNPs within each gene and the PGSC gene annotation.

https://doi.org/10.1371/journal.pone.0036347.s006

(XLSX)

Table S3.

Number of SNPs removed prior to mapping. A list of the categories of SNPs removed prior to mapping and the number of SNPs removed for each category.

https://doi.org/10.1371/journal.pone.0036347.s007

(XLSX)

Table S4.

Comparison of linkage group sizes between populations DRH and D84. Size differences between linkage groups in populations DRH and D84 demonstrating reduced recombination in D84.

https://doi.org/10.1371/journal.pone.0036347.s008

(XLSX)

Table S5.

List of disconcordant markers. A list of markers for which chromosome assignment based on the genetic maps differs from the chromosome assignment based on the PGSC pseudomolecules. Also included are marker ID numbers, segregation status, annotation and the possibility of a paralog for the marker (based on GenomeStudio graphs).

https://doi.org/10.1371/journal.pone.0036347.s009

(XLSX)

Table S6.

List of superscaffolds with mis-alignments. Superscaffolds that appear to have mis-alignments based on comparison with the genetic maps.

https://doi.org/10.1371/journal.pone.0036347.s010

(XLSX)

Acknowledgments

The authors wish to thank Dr. Daniel Zarka for genotyping both populations and Donna Kells for providing tissue culture plantlets for DNA extractions and mini-tuber development. The contributions of Suzanne Piovano and Tatiana Boluarte-Medina for development of the DRH population and DNA extraction are gratefully acknowledged.

Author Contributions

Conceived and designed the experiments: DSD REV CRB. Performed the experiments: KJF JJC. Analyzed the data: KJF JJC ANM JPH CNH. Contributed reagents/materials/analysis tools: REV CRB DSD. Wrote the paper: KJF JJC ANM CNH JPH REV CRB DSD.

Citation: Felcher KJ, Coombs JJ, Massa AN, Hansey CN, Hamilton JP, Veilleux RE, et al. (2012) Integration of Two Diploid Potato Linkage Maps with the Potato Genome Sequence. PLoS ONE 7(4): e36347. https://doi.org/10.1371/journal.pone.0036347

References

1. FAOSTAT (2007) FAOSTAT2007Available: http://faostat.fao.org/site/339/default.aspx via the Internet. Accessed 31 May 2010. Available: http://faostat.fao.org/site/339/default.aspx via the Internet. Accessed 31 May 2010.

2. Bonierbale MW, Plaisted RL, Tanksley SD (1988) RFLP maps based on a common set of clones reveal modes of chromosomal evolution in potato and tomato. Genetics 120: 1095–1103.MW BonierbaleRL PlaistedSD Tanksley1988RFLP maps based on a common set of clones reveal modes of chromosomal evolution in potato and tomato.Genetics12010951103

3. Gebhardt C, Ritter R, Barone A, Debener T, Walkemeier B, et al. (1991) RFLP maps of potato and their alignment with the homoeologous tomato genome. Theor Appl Genet 83: 49–57.C. GebhardtR. RitterA. BaroneT. DebenerB. Walkemeier1991RFLP maps of potato and their alignment with the homoeologous tomato genome.Theor Appl Genet834957

4. Jacobs JME, Van Eck HJ, Arens P, Verkerk-Bakker B, te Lintel Hekkert B, et al. (1995) A genetic map of potato (Solanum tuberosum) integrating molecular markers, including transposons, and classical markers. Theor Appl Genet 91: 289–300.JME JacobsHJ Van EckP. ArensB. Verkerk-BakkerB. te Lintel Hekkert1995A genetic map of potato (Solanum tuberosum) integrating molecular markers, including transposons, and classical markers.Theor Appl Genet91289300

5. Van Os H, Andrzejewski S, Bakker E, Barrena I, Bryan GJ, et al. (2006) Construction of a 10,000-marker ultradense genetic recombination map of potato: providing a framework for accelerated gene isolation and a genome-wide physical map. Genetics 173: 1075–1087.H. Van OsS. AndrzejewskiE. BakkerI. BarrenaGJ Bryan2006Construction of a 10,000-marker ultradense genetic recombination map of potato: providing a framework for accelerated gene isolation and a genome-wide physical map.Genetics17310751087

6. Caromel B, Mugniéry D, Lefebvre V, Andrzejewski S, Ellissèche D, et al. (2003) Mapping QTLs for resistance against Globodera pallida (Stone) Pa2/3 in a diploid potato progeny originating from Solanum spegazzinii. Theor Appl Genet 106: 1517–1523.B. CaromelD. MugniéryV. LefebvreS. AndrzejewskiD. Ellissèche2003Mapping QTLs for resistance against Globodera pallida (Stone) Pa2/3 in a diploid potato progeny originating from Solanum spegazzinii.Theor Appl Genet10615171523

7. Milbourne D, Meyer RC, Collins AJ, Ramsay LD, Gebhardt C, et al. (1998) Isolation, characterization and mapping of simple sequence repeat loci in potato. Mol Gen Genet 259: 233–245.D. MilbourneRC MeyerAJ CollinsLD RamsayC. Gebhardt1998Isolation, characterization and mapping of simple sequence repeat loci in potato.Mol Gen Genet259233245

8. Anithakumari AM, Tang J, van Eck HJ, Visser RGF, Leunissen JAM, et al. (2010) A pipeline for high throughput detection and mapping of SNPs from EST databases. Mol. Breeding 26: 65–75.AM AnithakumariJ. TangHJ van EckRGF VisserJAM Leunissen2010A pipeline for high throughput detection and mapping of SNPs from EST databases. Mol.Breeding266575

9. Ching A, Caldwell K, Jung M, Dolan M, Smith O, et al. (2002) SNP frequency, haplotype structure and linkage disequilibrium in elite maize inbred lines. BMC Genetics 3: 19.A. ChingK. CaldwellM. JungM. DolanO. Smith2002SNP frequency, haplotype structure and linkage disequilibrium in elite maize inbred lines.BMC Genetics319

10. Zhu YL, Song QJ, Hyten DL, Van Tassell CP, Matukumalli LK, et al. (2003) Single-nucleotide polymorphisms in soybean. Genetics 163: 1123–1134.YL ZhuQJ SongDL HytenCP Van TassellLK Matukumalli2003Single-nucleotide polymorphisms in soybean.Genetics16311231134

11. Shen YJ, Jiang H, Jin JP, Zhang ZB, Xi B, et al. (2004) Development of genome-wide DNA polymorphism database for map-based cloning of rice genes. Plant Physiol 135: 1198–1205.YJ ShenH. JiangJP JinZB ZhangB. Xi2004Development of genome-wide DNA polymorphism database for map-based cloning of rice genes.Plant Physiol13511981205

12. International Rice Genome Sequencing Project (2005) The map-based sequence of the rice genome. Nature 436: 793–800.International Rice Genome Sequencing Project2005The map-based sequence of the rice genome.Nature436793800

13. Schnable PS, Ware D, Fulton RS, Stein JC, Wei F, et al. (2009) The B73 maize genome: complexity, diversity, and dynamics. Science 326: 1112–1115.PS SchnableD. WareRS FultonJC SteinF. Wei2009The B73 maize genome: complexity, diversity, and dynamics.Science32611121115

14. Schmutz J, Cannon SB, Schlueter J, Ma J, Mitros T, et al. (2010) Genome sequence of the palaeopolyploid soybean. Nature 463: 178–183.J. SchmutzSB CannonJ. SchlueterJ. MaT. Mitros2010Genome sequence of the palaeopolyploid soybean.Nature463178183

15. Potato Genome Sequencing Consortium (2011) Genome sequence and analysis of the tuber crop potato. Nature 475: 186–195.Potato Genome Sequencing Consortium2011Genome sequence and analysis of the tuber crop potato.Nature475186195

16. McNally KL, Childs KL, Bohnert R, Davidson RM, Zhao K, et al. (2009) Genomewide SNP variation reveals relationships among landraces and modern varieties of rice. Proc Natl Acad Sci 106: 12273–12278.KL McNallyKL ChildsR. BohnertRM DavidsonK. Zhao2009Genomewide SNP variation reveals relationships among landraces and modern varieties of rice.Proc Natl Acad Sci1061227312278

17. Gore MA, Chia JM, Elshire RJ, Sun Q, Ersoz ES, et al. (2009) A first-generation haplotype map of maize. Science 326: 1115–1117.MA GoreJM ChiaRJ ElshireQ. SunES Ersoz2009A first-generation haplotype map of maize.Science32611151117

18. Hamilton JP, Hansey CN, Whitty BR, Stoffel K, Massa AN, et al. (2011) Single nucleotide polymorphism discovery in elite North American potato germplasm. BMC Genomics 12: 302.JP HamiltonCN HanseyBR WhittyK. StoffelAN Massa2011Single nucleotide polymorphism discovery in elite North American potato germplasm.BMC Genomics12302

19. Solanaceae Cooperative Agricultural Project (2011) Solanaceae Cooperative Agricultural Project2011Available: http://solcap.msu.edu/ via the Internet. Accessed 11 Nov 2011. Available: http://solcap.msu.edu/ via the Internet. Accessed 11 Nov 2011.

20. Paz MM, Veilleux RE (1999) Influence of culture medium and in vitro conditions on shoot regeneration in Solanum phureja monoploids and fertility of regenerated doubled monoploids. Plant Breed 118: 53–57.MM PazRE Veilleux1999Influence of culture medium and in vitro conditions on shoot regeneration in Solanum phureja monoploids and fertility of regenerated doubled monoploids.Plant Breed1185357

21. Douches DS, Quiros CF (1988) Genetic recombination in a diploid synaptic mutant and a Solanum tuberosum×S. chacoense diploid hybrid. Heredity 60: 183–191.DS DouchesCF Quiros1988Genetic recombination in a diploid synaptic mutant and a Solanum tuberosum×S. chacoense diploid hybrid.Heredity60183191

22. Rouppe vander Voort JNAM, van Zandvoort P, van Eck HJ, Folkertsma RT, Hutten RCB, et al. (1997) Use of allele specificity of co-migrating AFLP markers to align genetic maps from different potato genotypes. Mol Gen Genet 255: 428–447.JNAM Rouppe vander VoortP. van ZandvoortHJ van EckRT FolkertsmaRCB Hutten1997Use of allele specificity of co-migrating AFLP markers to align genetic maps from different potato genotypes.Mol Gen Genet255428447

23. Veilleux RE, Miller AR (1998) Hybrid breakdown in the F1, between Solanum chacoense and S. phureja and gene transfer for leptine biosynthesis. J Amer Soc Hort Sci 123: 854–858.RE VeilleuxAR Miller1998Hybrid breakdown in the F1, between Solanum chacoense and S. phureja and gene transfer for leptine biosynthesis.J Amer Soc Hort Sci123854858

24. Birhman RK, Laublin G, Cappadocia M (1994) Inheritance of a lethal yellow-cotyledon seedling mutant in Solanum chacoense Bitt. J Hered 85: 241–242.RK BirhmanG. LaublinM. Cappadocia1994Inheritance of a lethal yellow-cotyledon seedling mutant in Solanum chacoense Bitt.J Hered85241242

25. Hu TT, Pattyn P, Bakker EG, Cao J, Cheng J, et al. (2011) The Arabidopsis lyrata genome sequence and the basis of rapid genome size change. Nature Genet 43: 476–481.TT HuP. PattynEG BakkerJ. CaoJ. Cheng2011The Arabidopsis lyrata genome sequence and the basis of rapid genome size change.Nature Genet43476481

26. JoinMap4 (2006) JoinMap42006Available: http://www.kyazma.nl/index.php/mc.JoinMap via the Internet. Accessed 11 Nov 2011. Available: http://www.kyazma.nl/index.php/mc.JoinMap via the Internet. Accessed 11 Nov 2011.

27. Slater GS, Birney E (2005) Automated generation of heuristics for biological sequence comparison. BMC Bioinformatics 6: 31.GS SlaterE. Birney2005Automated generation of heuristics for biological sequence comparison.BMC Bioinformatics631

28. Rezvoy CM, Charif D, Guéguen L, Marais GAB (2007) MareyMap: an R-based tool with graphical interface for estimating recombination rates. Bioinformatics 23: 2188–2189.CM RezvoyD. CharifL. GuéguenGAB Marais2007MareyMap: an R-based tool with graphical interface for estimating recombination rates.Bioinformatics2321882189