Description of the maritime pine unigene set

We obtained 2,017,226 high-quality sequences, 1,892,684 of which belonged to the 73,883 multisequence clusters (or contigs) identified, the remaining 124,542 ESTs corresponding to singletons. This created a gene index of 198,425 different sequences, assuming that the singleton ESTs corresponded to unique transcripts. The number of unique sequences is almost certainly overestimated, because some sequences probably arise from non-overlapping regions of the same cDNA or correspond to alternative transcripts. The assembly was denoted PineContig_v2 and is available from [22].

SNP-assay genotyping statistics

We used the maritime pine unigene set to develop a 12 k SNP array for use in genetic linkage mapping. The mean call rate (percentage of valid genotype calls) was 91% and 94% for the G2 and F2 mapping populations, respectively.

In the G2 pedigree, we found 2,264 polymorphic loci (2,210 SNPs and 54 indels) corresponding to 1,473 PineContig_v2 contigs, including 1,660 SNPs segregating in a 1:1 ratio (760 and 900 SNPs being informative for the female and the male parent, respectively) and 604 SNPs segregating in a 1:2:1 ratio.

In the F2 pedigree, we found 1,215 polymorphic loci (1,184 SNPs and 31 indels) segregating in a 1:2:1 ratio and corresponding to 881 PineContig_v2 contigs.

The conversion rate (number of polymorphic SNPs/indels divided by the total number of SNPs/indels in the assay, that is, 9,279 SNPs) was 24.4% for the G2 population and 13.1% for the F2 population. The conversion rates for SNPs resulting in nucleotide replacement were 35% for the G2 population and 18% for the F2 population, whereas those for 1 bp indel mutations were almost zero (1.8% for G2 and 1% for F2). Indels should, therefore, be avoided when designing an Infinium assay on the basis of 454 reads. Polymorphic SNPs were made available through the National Center for Biotechnology Information (NCBI) dbSNP database [23]. The accession numbers are listed in Additional file 2.

Validation of the SNP assay

The presence of several SNPs within a single contig made it possible to validate the genotyping assay. For the F2 population, 215 contigs contained more than one SNP. We carried out 22,712 genotyping comparisons and found no genotyping inconsistencies between SNPs from the same contig. Thus, assuming that the probability of crossover between SNPs from the same contig is zero between generations, we obtained a genotyping error of 0%. For the G2 population, 424 contigs contained more than one SNP. We carried out 91,015 genotype comparisons and detected 154 recombination events between SNPs from the same contig, corresponding to a genotyping error of 0.17%. This result confirms the high reproducibility of customized Infinium assays based on careful bioinformatic analysis.

Comparison of segregation patterns between inbred and outbred mating types

We searched for chromosomal regions containing gene loci with sublethal or lethal alleles, by comparing the segregation pattern for the offspring obtained by outbreeding (G2 pedigree) with that for the offspring obtained by inbreeding (F2 pedigree). This test was based on the assumption that clusters of distorted SNP markers in the F2 progeny that are not distorted in the G2 progeny indicate the presence of lethal or sublethal equivalents revealed by inbreeding. Using a type I error risk of 1%, we initially found 27 distorted markers in the F2 progeny, 25 of which were clustered at three locations: two in linkage group (LG) 10 and one in LG2 [see Additional file 3 and Figure 2]. The number of offspring genotyped in the F2 progeny differed between SNP assays (from 69 F2s for the 12 k SNP assay to 472 F2s for the 384-plex assay) and the segregation data were dependent on sample size. We, therefore, systematically checked clustered distorted SNP markers from the 12 k assay in a larger sample (380 F2s) with a targeted medium-throughput genotyping assay (Mass Array, Sequenom). Only one cluster of four distorted markers remained in LG2 after this validation step [see Additional file 4]. Only two distorted markers gave Basic Local Alignment Search Tool (BLAST) X hits in SwissProt (Probable histone H2A.3 for ‘SNPnew127’ and ‘SNPnew128’ in contig F51TW9001A6567 of PineContigv2). No BLAST hit was found for loci ‘m306’ in contig CL2488CT12CN14 and ‘SNPnew25’ in contig BX254626. A detailed analysis of allele transmission from the grandparental genotypes to the F2 offspring clearly showed that the sublethal allele was inherited from the Corsican paternal grandparent [see Additional file 5].

We also detected segregation distortion in the G2 pedigree for 13 (2.4%) and 6 (1%) markers in the G2F and G2M maps, respectively. One group of eight distorted markers clustered in LG6 of the G2F map, whereas the other distorted loci were distributed more or less evenly, in all LGs. However, it was not possible to verify this cluster of distorted markers in a larger sample, because of the small number of genotypes available for the G2 pedigree. The genetic basis of the segregation distortion may be the abortion of male or female gametes (prezygotic selection) or the selective fertilization of particular genotypes (postzygotic selection). In the case of prezygotic selection, we would expect to see segregation distortion in one of the parental maps only, whereas we would expect to see segregation distortion in the corresponding genomic region on both parental maps for postzygotic selection. If real, given that this hotspot of distorted markers was found only on the G2F map, it could indicate the presence of a locus under female gametic selection.

Mapping results

Factors affecting recombination

We used the Wilcoxon signed rank test to test the hypothesis that ‘map lengths are equal between the three maps: G2F, G2M and F2.’ This hypothesis was not rejected for the comparison between G2F and G2M, P-value _{(G2F versus G2M)} = 0.78, indicating that sex had no significant effect on map length in this mapping population. The same test was applied for the comparisons between G2F and F2 and between G2M and F2, with significant differences detected in both cases: P-value _{(F2 versus G2F)} = 0.0004 and P-value _{(F2 versus G2M)} = 0.005. We checked that the effect of genetic background on the frequency of recombination was not due to the presence of more markers on the F2 map than on the G2F and G2M maps, by carrying out a Wilcoxon signed rank test for all pair-wise recombinations between the common markers in each LG. This test clearly showed that the ‘genetic status’ (intra- versus interprovenance hybrids in our case) of the parental genotypes in which meiotic recombination occurred had a significant effect on the frequency of recombination, with nine LGs presenting significant differences between both F2 and G2F, and F2 and G2M, three LGs presenting a significant difference between F2 and G2F or F2 and G2M, and one LG presenting no effect [see Additional file 8]. Finally, a Z-test was applied to each pair-wise comparison, for the identification of significant pairs among those used to perform the Wilcoxon rank-test (highlighted in red in Figure 3). There was a clear trend toward a greater incidence of significant pair-wise recombination for the F2 map (interprovenance hybrid) than for the G2F or G2M maps (intraprovenance hybrids).

Distribution of recombination along the chromosomes

We also investigated whether the distribution of recombination along the maritime pine chromosomes was affected by the genetic background in which meiotic recombination occurred, by kernel density function analysis. This approach made it possible to set appropriate band widths (per map and per LG) for gene counts, rather than having to fix an arbitrary interval, as in most methods. Based on a comparative analysis of observed and expected marker distributions, we first determined the upper and lower thresholds defining recombination hotspots (larger gaps between markers than expected and coldspots (tightly linked markers), respectively [see Additional file 9]. An analysis of the F2 map showed that a cluster of at least 10 markers (P = 3 × 10^-9) could be considered to constitute a recombination coldspot, whereas a cluster of no more than three markers (P = 3.6 × 10^-10) could be interpreted as a recombination hotspot. For the G2F and G2M maps, recombination coldspots were defined as a cluster of at least eight markers (P _G2F = 0.002; P _G2M = 4.5 × 10^-25), whereas hotspots were defined as a cluster of no more than two markers (P _G2F = 0.002; P_G2M = 1.4 × 10^-26). A plot of gene density over each linkage group, generated by sliding (every 1 cM) an interval corresponding to the predetermined bandwidth, revealed the presence of significant gene clusters or gaps in the three maps (Figure 4 and Additional file 10). By aligning homologous linkage groups, we were able to compare the numbers and locations of recombination coldspots and hotspots between the three maps obtained for the different genotypes (two intraprovenance hybrids for the G2 population and one interprovenance hybrid for the F2 population). We detected a mean of 2.8 coldspots and 5.6 hotspots of recombination per chromosome, respectively. Most (67%) of the hotspots were common to at least two genotypes (27% being common to all three genotypes), but only 48% of the coldspots were common to at least two genotypes (only 7.5% were common to all three genotypes). This result suggests that the spatial structure of recombination is genetically variable, with some recombination hotspots and coldspots specific to a given genotype. Based on the number of shared and specific recombination coldspots and hotspots (Venn diagram in Additional file 10), we calculated a Jaccard index to assess the similarity between the three maps (three pair-wise comparisons). Surprisingly, the recombination patterns of the G2F and G2M maps were found to be more similar to that of the F2 map than to each other.

Development of genomic tools to facilitate genetic research in maritime pine

Comparison of segregation patterns between inbred and outbred matings indicates the presence of a chromosomal region with a deleterious mutation acting at the postzygotic stage

Departure from Mendelian expectations, which is also known as segregation distortion (SD, [38]), is frequently reported in linkage mapping studies (reviewed by Li et al. [39]). If a gene causing SD is segregating in a population, then the markers close to it tend to display distorted segregation ratios. Thus, as a rule of thumb, the clustering of markers displaying SD in particular genomic regions (so-called segregation distortion regions, SDRs) may indicate that segregation distortion is caused by genetic factors rather than statistical bias or genotyping errors. However, as illustrated in this study, small population size may lead to false positives and the identification of spurious SDRs. Care should therefore be taken to validate SDRs before any biological interpretation is attempted.

Biologically, aberrant Mendelian segregation can be attributed to selection occurring at different stages of the plant’s life cycle, from gametophyte development to seed germination and plant growth [20, 40]. In this study, a single cluster of distorted markers was detected and validated in LG2 of the F2 map, whereas the corresponding genomic region on the two G2 maps displayed no deviation from the expected Mendelian segregation ratio. This strongly suggests the presence of a deleterious mutation (or a cluster of tightly linked embryo viability loci), revealed by inbreeding, that influences the fitness of the F2 zygotes at some point between fertilization and the age of 10 years (as the tissues sampled for DNA extraction were taken from 10-year-old trees). This conclusion is supported by two additional observations. First, this F2 family was selected specifically because of its low rate of seed abortion (frequency of embryo-less seeds, as estimated by assessing floating in water, was lower than for other available F2s from the maritime pine breeding program, unpublished results), making it particularly suitable to genetic analysis requiring a large sample size. In our study, 638 seeds were initially planted in a nursery in June 1998; 626 seedlings germinated (that is, only 1.9% died soon after germination) and were transplanted into the field in March 1999. Total height was then measured every fall, beginning in 1999. Fifteen seedlings died during the first growing season in the field (assessment in the fall of 1999). The following year, 43 other seedlings died, but no further deaths were recorded thereafter. It is difficult to determine whether these deaths were due to some crisis during transplantation from the nursery to the field or to genetic load. However, peak mortality did not occur in the nursery or just after field transfer, and the semilethal allele was inherited from the Corsican paternal grandparent. These findings suggest that this SDR decreases the fitness of homozygous Corsican genotypes in early stages of development and later in tree growth. Unfortunately, no post-mortem analysis involving the sampling of plant material from the whole progeny just after germination was performed, to determine whether the dead plants were all homozygous for the Corsican allele in the SDR concerned.

Second, in a previous study, Plomion et al. [41] compared the segregation patterns of random amplified polymorphic DNA (RAPD) markers in megagametophytes (a maternally derived haploid tissue surrounding the embryo) from the same hybrid tree (H12), sampled from either inbred (self-cross) or outbred (open-pollinated cross) seeds. They observed no significant SD for loci in the dataset resulting from selfing, suggesting that gametic selection, leading to gamete abortion or lower gamete fitness, can be ruled out as a possible cause of SD in this study.

Genomic regions containing lethal or sublethal alleles have already been detected in several conifers, through linkage mapping approaches (reviewed in Williams [21]). The number of such lethal or sublethal equivalents is generally high in populations, as revealed by the typical high level of inbreeding depression in these outcrossing species (reviewed by Williams and Savolainen [42]), but their severity varies in the population, with some genotypes (like that selected in this study) bearing mutations that are, a priori, less deleterious than others. The nature of the underlying loci remains unclear. Some of these genetic factors are involved in early embryo development, resulting in a lower yield of filled seeds upon selfing, others decrease seedling growth and cause abnormal phenotypes, whereas others are directly involved in seedling mortality at later stages of development, from a few weeks [43] to a few months after germination, as shown here. In addition to providing fundamental knowledge, the analysis of segregation distortion and the identification of SDR are of great importance for the correct determination of quantitative trait loci (QTL) positions and for the estimation of QTL effects. Indeed, SD influences the estimation of recombination frequency and may, therefore, decrease the accuracy of QTL mapping in this mapping population.

The extent and spatial distribution of meiotic recombination is genetically variable

Recombination is a driving force behind the generation of genetic diversity and is also a key process shaping genomic architecture [19]. An understanding of the factors controlling the frequency and genomic distribution of meiotic recombination is, therefore, essential if we are to manipulate this process to improve breeding accuracy. This study generated three major results.

First, we confirmed that, despite their large physical size, pine chromosomes display a similar number of crossover events to other smaller plant chromosomes. This observation led Thurieaux [44] to suggest that recombination was confined largely to the coding regions, because all eukaryotes have approximately the same number of genes, as demonstrated by the genome sequences of various organisms (for example, in Arabidopsis [45], rice [46, 47], maize [48] and sorghum [49]), although other genomic features may affect recombination. At the microscale, no consistent relationship has yet been established between recombination rate and gene content [50, 51], suggesting that it is probably not correct to assume that all plant recombination hotspots correspond to gene-rich regions. It will not be possible to determine whether recombination hotspots correspond to gene-rich regions in conifers until a complete conifer genome sequence is obtained. However, as gene-rich regions tend to be associated with high rates of recombination in other plants, it seems likely that relationships between crossover frequency and gene density will not deviate from this trend in conifers. For example, in bread wheat (17Gb/C), a non-uniform crossover gradient along chromosome 3B has been observed, with lower frequencies of crossover in the gene-poor centromeric region and the highest frequencies of crossover in the distal subtelomeric regions, in which gene density is higher [52]. At a finer scale, these authors also demonstrated that gene content was one of the factors driving recombination in this species [53].

Second, we observed that meiotic recombination was not randomly distributed along the length of the maritime pine chromosomes, suggesting that recombination occurs at specific sites, the recombination hotspots (reviewed by Lichten and Goldman [54]). An uneven distribution of markers is a classical observation in most papers reporting saturated linkage maps for plants and animals. Tests of departure from a Poisson distribution have always been based on a single or a series of different, arbitrarily fixed intervals (as illustrated by Moriguchi et al. [36] in Cryptomeria japonica). To our knowledge, only Pavy et al. [8] have previously implemented a statistical approach, based on kernel density function, in Picea spp., to overcome the need to use such fixed bandwidths in analyses of ‘gene-rich regions’ as an indicator of suppressed recombination. In this study, we used the same strategy, combining it with a sliding window approach, to improve the resolution of recombination hotspots and coldspots. Interestingly, in most LGs of the G2F and G2M maps, a sharp cold spot located in the middle of the linkage group was surrounded by two large hot spots. This suggests that these cold spots may correspond to the centromeric regions of the chromosome, in which the frequency of recombination is known to be low [51–55] and in which markers tend to cluster on meiotic maps. However, further studies are required to confirm this assertion. This signature was less clear in the F2 map, which contained about twice as many coldspots as the G2 maps (48 in F2 versus 27 in G2F and 28 in G2M), with a similar number of hotspots (71 versus 62 and 69). An uneven distribution of crossover events has been reported for both species with small genomes and those with large genomes ([51] for Arabidopsis, [53] for wheat) and an understanding of the distribution of recombination events is critical for various genetic applications. First, following on from the discussion above, if recombination occurs in hotspots and these hotspots bear most of the genes, then differential sequencing efforts will be required to obtain data for all of the genes in conifer genome sequencing programs. Second, as illustrated by Wang et al. for rice [56], the map-based cloning of a QTL is facilitated if the QTL is located in a genomic region containing a recombination hotspot, simply because it is easier to identify large numbers of recombinants from segregating populations. This information may be useful for the characterization of genes underlying major QTLs in species with large genomes, such as pines, as already reported for wheat [57].

Third, our results show that the extent and spatial distribution of meiotic recombination is genetically variable. The interprovenance hybrid had recombination rates 1.2 times higher (measured on the basis of total map distance) than those of either of the intraprovenance hybrids. This suggests that the genetic divergence of bivalents may account for the extent of recombination at meiosis. However, a comparison of gene heterozygozity between the three genotypes on the basis of both mapping data [see Additional file 6] and the in silico prediction of polymorphisms [see Additional file 12] showed that the diversity of the interprovenance hybrid was intermediate with respect to the diversity of the two intraprovenance hybrids. These two findings indicate that the genetic distance (at least within the gene space, in which most crossover events are thought to occur) between the bivalents does not alter meiotic pairing to a point that would lead to differences in recombination frequencies, as shown in interspecific hybrids by in situ hybridization [58] and linkage mapping [59]. Moreover, the high degree of collinearity between the maps for the intra- and interprovenance hybrids shows that no genome rearrangement occurred during hybridization that might have led to a recombination disorder. We can conclude that the observed difference in map length reflects differences between genotypes. The distribution of recombination events differed between the three genotypes, which had only some hotspots, and even fewer coldspots in common. This suggests that the spatial pattern of recombination along the chromosome is also genetically variable and under polygenic control, as demonstrated by Comeron et al. [16] in Drosophila melanogaster. Recombination is known to be genetically variable [15, 60, 61] and under the control of multiple trans and cis genetic modifiers. Sequence polymorphisms [62, 63] and/or the methylation status of these genetic factors may underlie these differences in recombination pattern and should be investigated further in conifers.

Whether the results obtained depend on the type of markers used needs to be addressed. First, it should be noticed that the total map length obtained in the present study with coding sequences, was similar to that obtained for the same genotypes using anonymous RAPD [41] or amplified fragment length polymorphism (AFLP) [3] markers (supposedly corresponding to non-coding DNA). Second, maps combining gene-based markers and genomic DNA markers (for example, proteins and RAPDs in [64], EST-Ps and AFLPs in [37], SNPs and AFLPs in [25]) were also constructed in this species and did not show any clustering of one or another marker type. Therefore, it is assumed that the recombinational landscape presented in this paper should not be biased by the type of marker (coding versus non-coding) used for linkage analysis.

Genetic material and DNA extraction

The two mapping populations (G2 and F2, Additional file 13) used in this study have been described elsewhere [25]: G2 is a three-generation outbred pedigree (full-sib progeny), whereas F2 is a three-generation inbred pedigree. Young needles of each individual were harvested and stored at −80°C until DNA extraction. Pieces of frozen needles (around 30 to 40 mg) were crushed using a mixer mill (Retsch MM300, Haan, Germany). Genomic DNA was isolated with the Invisorb Plant DNA 96 kit from Invitek GmbH (Berlin, Germany), according to the manufacturer’s instructions. All concentrations were determined with a Nanodrop spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA) and fluorescence assays (Quant-IT kit, Invitrogen, Carlsbad, CA, USA). All samples with concentrations exceeding 50 ng/μl (based on fluorescence measurements) were kept and sent to Genediffusion (Pasteur Institute, Lille, France) for Infinium assays on the iScan platform (Illumina Inc., San Diego, CA, USA).

Bioinformatic analysis

Detection of nucleotide polymorphism

Four subsets of this vast body of data (detailed below) were screened for the development of the 12 k Illumina Infinium SNP array. A flowchart describing the steps involved in the identification of SNPs segregating in the Aquitaine population is shown in Figure 5.

•  In silico SNPs detected in Aquitaine genotypes (set#1). In total, 685,926 sequences from Aquitaine genotypes (454 and Sanger reads) derived from 17 cDNA libraries were extracted from PineContig_v2 [see Additional file 15]. We focused on this ecotype of maritime pine because our long-term objective is to carry out genomic selection in the breeding program focusing principally on this provenance. Data were cleaned with the SmartKitCleaner and Pyrocleaner tools [65]. The remaining 584,089 reads were distributed into 42,682 contigs (10,830 singletons, 15,807 contigs with 2 to 4 reads, 6,871 contigs with 5 to 10 reads, 3,927 contigs with 11 to 20 reads, 5,247 contigs with more than 20 reads, Additional file 16). SNP detection was performed for contigs containing more than 10 reads. A first Perl script (‘mask’) was used to mask singleton SNPs [71]. A second Perl script, ‘Remove’, was then used to remove the positions containing alignment gaps for all reads. The number of false positives was minimized by establishing a priority list of SNPs in the assay on the basis of MAF, depending on the depth of each SNP. Finally, a third script, ‘snp2illumina’, was used to extract SNPs and short indels of less than 7 bp, which were output as a SequenceList file compatible with Illumina ADT software. The resulting file contained the SNP names and surrounding sequences with polymorphic loci indicated by IUPAC codes for degenerate bases. We generated statistical data for each SNP — MAF, minimum allele number (MAN), depth and frequencies of each nucleotide for a given SNP — with a fourth script, ‘SNP_statistics’. We established the final set of SNPs by considering as ‘true’ (that is, not due to sequencing errors) all non-singleton biallelic polymorphisms detected on more than five reads, with a MAF of at least 33% and an Illumina score greater than 0.75 (Filter 2 in Figure 5). Based on these filter parameters, 10,224 polymorphisms (SNPs and 1 bp insertion/deletions, referred to hereafter as SNPs) were detected

•  In silico SNPs detected in comparative orthologous sequences (COS) between pine and spruce (set#2). For the COS between pine and spruce identified by Chancerel et al. [25], 303 resulted in a hit with at least one maritime pine EST. By blasting these sequences against PineContig_v2 (BlastN, Evalue = 10^-4), we identified 292 contigs containing 185 COS markers, 53 of which were already present in set#1, and 132 of which were specific to set#2.

•  In vitro SNPs selected from a previous SNP array (set#3). In total, 223 SNPs originally detected and validated in a set of 41 resequenced candidate genes (in vitro SNPs [71]) were selected for the 12 k SNP assay.

•  SNPs detected in a gene fragment including at least one SNP associated with growth (set#4). An SNP associated with height growth has been identified [72]) in contig CT-3782 of the first Pinus pinaster assembly described by Chancerel et al. [25]. By blasting the CT-3782 contig sequence against PineContig_v2 (BlastN, Evalue = 10^-4), we identified BX248795 as homologous to CT-3782. We found 14 in silico SNPs in this contig, which were included in the 12 k SNP assay.

Finally, based on these four different SNP sets, 10,593 SNPs (corresponding to 11,938 bead types, Additional file 17) were available for genotyping after filtering with the ADT of Illumina. All but three of the SNPs had a score above 0.63.

SNP genotyping assay

Genotyping was carried out at Genediffusion (Institut Pasteur, Lille, France) with the Illumina Infinium assay, used according to the manufacturer’s instructions (Illumina). In total, 87 and 70 offspring were initially genotyped for the G2 and F2 mapping populations, respectively. The Infinium assay is based on the direct hybridization of genomic targets to array-bound sequences. Single-base extension is followed by fluorescence staining, signal amplification, scanning and analysis with Genome Studio software v. 1.0. From the initial set of 10,593 SNPs, 1,314 (12.4%) did not pass Illumina production quality control and were eliminated. The remaining 9,279 SNPs (6,299 SNPs sensu stricto and 2,980 indels distributed in 4,613 different contigs) were individually inspected with Genome Studio software, with a GenCall score cutoff of 0.15 (according to Illumina’s recommendations) to detect failed, monomorphic and polymorphic SNPs. We considered loci for which two or three scatter plots (depending on the type of marker segregation) were identified without ambiguity to be polymorphic markers. SNP clusters were modified manually, to refine cluster positions when necessary.

Analysis of segregation distortion

For each locus, we tested the pattern of allelic segregation for goodness-of-fit to expected Mendelian segregation ratios, in Chi² tests (P ≤0.01). We assumed that clusters of distorted loci in the F2 inbred progeny that were not distorted in the G2 outbred progeny indicated the presence of a deleterious allele revealed by inbreeding. Given the small number of F2 progenies (69) genotyped with the 12 k SNP array, markers displaying SD were examined on a larger and independent sample of F2s (380 trees), to check for the presence of hotspots of SD. We used the medium-throughput MassARRAY iPLEX genotyping assay from Sequenom (San Diego, CA, USA) for this purpose.

In total, 34 SNPs — 25 distorted (in 19 contigs) and 9 non distorted markers (in 9 contigs distributed in different LGs) — in the F2 progeny were included in two multiplex assays (22plex and 12plex, Additional file 18) with MassArray assay design 4.1 software (Sequenom). Six pairs of SNPs displaying SD and located in the same contig were used in the assay, to assess the reproducibility of this genotyping method. Four pairs were successfully genotyped and showed no genotyping inconsistencies. The hybrid parent (H12) used as a positive control also displayed no genotyping inconsistencies, confirming the high degree of reproducibility of the iPLEX GOLD method.

DNA extraction and quantification were carried out as described above. In total, 15 ng of DNA was required for the reaction. Genotyping was carried out at the Genomic and Sequencing Facility of Bordeaux (France), with the iPLEX Gold genotyping kit (Sequenom), according to the manufacturer’s instructions. The iPLEX Gold SNP genotyping method involves several steps: PCR amplification is carried out first, followed by SAP treatment (to digest unincorporated dNTPs). A single-base extension reaction is then performed, followed by an ion-exchange cleanup step. Finally, the products are detected in a MassArray mass spectrophotometer and the data are acquired in real time with MassArray RT software. Alleles were automatically assigned by MassArray TyperAnalyser 4.0.22 software and associated with a reliability value. Positive (hybrid parent of the F2 mapping population) and negative controls were including in the genotyping process. Visual inspection was carried out for all the SNPs, to detect any incorrect assignments made by the ‘Autocluster’ option of the MassArray Typer Analyser software. Finally, locus segregation was tested for goodness of fit to expected Mendelian segregation ratios, in Chi² tests (P ≤0.01).

Linkage mapping strategy

For linkage analysis, we retained only one SNP if several were present with the same contig.

Estimation of genome length and map coverage

where R is the haploid number of chromosomes (12 in our case), N is the number of loci positioned on map 1, X is the maximum observed map distance between two adjacent markers in cM, at or above a minimum LOD threshold value of 6, 7 and 8. X was set to 50 cM for the F2 maps, to take into account the splitting of LG8 into two subgroups.

Analysis of marker distribution and comparison of recombination frequencies

Availability

The browsing of maritime pine contigs, annotations and SNPs and data mining by BioMart can be carried out at [22].

Information about the linkage maps is available from the PinusMap database, available from [80] and the Pine Cmap database, available from [81].

Roche 454 sequencing data are available at the short-read archive of the NCBI database [82].

Polymorphic SNPs are available from the NCBI dbSNP database [23]. Accession numbers are listed in Additional file 2.