Plant materials and DNA extraction
A yellow seeded B. rapa variety 'Yellow Sarson' and a Chinese cabbage doubled haploid (DH) line 'RI16' were crossed to develop a recombinant inbred (RI) line mapping population. The F1 plants were selfed to obtain F2 seeds and over 500 F2 plants were produced. From F2, the single seed descent (SSD) method was used to produce the following generations. Ninety-two F7 RI lines and the parental lines were selected for Illumina's Solexa sequencing and genetic map construction (Additional file 1). The DNA was extracted from leaves with a CTAB method as described previously .
Three sets of SRAP primers (Additional file 2) were included to produce SRAP products. The SRAP PCR running program was the same as described by Li and Quiros  and SRAP PCR was set up as described by Sun et al. . SRAP PCR reactions were performed in a 10 μl mixture containing 50 ng of genomic DNA, 375 μM dNTP, 0.15 μM of each primer, 1× PCR buffer, 1.5 mM MgCl2 and 1 unit of Taq polymerase. The SRAP PCR running program was 94°C for 3 min, 5 cycles of 94°C for 1.0 min, 35°C for 1.0 min, 72°C for 1.0 min, followed by 30 cycles of 94°C for 50 sec, 50°C for 50 sec, 72°C for 50 sec and final extension 72°C for 10 min.
The first set of SRAP primers consisted of fourteen forward primers (Set1F01-Set1F14) and 192 reverse primers (Set1R001-Set1R192), most of which were previously used in the construction of an ultradense genetic map in B. napus. Only the DNA of two parental lines was amplified with these primers in 14 384-well plates. In each 384-well plate, 2 DNA samples, one forward primer and 192 reverse primers were included to produce 384 PCR reactions. In total, 5,376 SRAP PCR reactions were set up with 2,688 SRAP primer pairs in 14 384-well plates.
The second set of SRAP primers consisted of four forward primers (Set2F01-Set2F04) and 96 reverse primers (Set2R01-Set2R96). These forward primers shared 12 identical nucleotides (5'-GAGTCCAAACCG-3') at the 5' end while all 96 reverse primers had another 12 nucleotides (5'-CGCAAGACCCAC-3') at the 5' ends. These four forward primers were combined with these 96 reverse primers to form 384 SRAP primer pairs. Ninety-four DNA samples were amplified with these 384 primer pairs in 94 384-well plates respectively.
The third set of SRAP primers consisted of one forward primer (Set3F01) and 384 reverse primers (Set3R001-Set3R384) that were combined to obtain 384 SRAP primer pairs. The forward primer was used previously and the reverse primers were selected from the primer collections used for gene cloning in the lab. The same PCR set-up was performed as in the second set of SRAP primers. Similarly, 94 384-well plates were used to amplify all ninety-four DNA samples.
Tagging of SRAP PCR products
The SRAP products from the first round of PCR were tagged in the second round of PCR with tagging primers.
To tag the SRAP PCR products obtained with the first set of SRAP primers, two 12 nucleotide heads (5'-GAGTCCAAACCG-3' and 5'-CGCAAGACCCAC-3') were joined respectively with the tails that consist of 11~14 nucleotides located at the 5' ends of the 14 forward SRAP primers to form tagging primers Set1T01-Set1T28 (Additional file 2). These 14 pairs of tagging primers were used to tag SRAP PCR products amplified with the two parental lines respectively, which allowed recognizing Solexa sequence originality. All tagging primers were used as forward primers and combined with the same 192 SRAP reverse primers (Set1R001-Set1R192) to amplify the SRAP PCR products in a specific PCR program.
To tag the SRAP PCR products obtained with the second set of SRAP primers, ninety-four tagging primers Set2T01-Set2T94 (Additional file 2) were designed for the 94 384-well plates in SRAP PCR. Since four forward primers of the second SRAP primers set had 12 identical nucleotides (5'-GAGTCCAAACCG-3') at the 5' ends, these nucleotides were used as tails and combined with ninety-four nine nucleotide heads to form ninety-four tagging primers. The tagging primers were used as forward primers. A 12 nucleotide (5'-CGCAAGACCCAC-3') primer that was shared by all ninety-four SRAP reverse primers in the second set of SRAP primers was used as reverse primer.
To tag the SRAP PCR products amplified with the third set of SRAP primers, another ninety four tagging primers were designed for the ninety-four 384-well plates in SRAP PCR. The first 12 nucleotides (5'-CGCAAGACCCAC-3') at the 5' end of the only SRAP forward primer was used as common 3' tails of tagging primers and ninety four different nine-nucleotide heads were added at the 5' end of the only SRAP forward primer to form ninety-four tagging primers (Set3T01-Set3T94). These tagging primers were used as forward primers and combined with the same 384 SRAP reverse primers.
The tagging PCR reactions were performed in a 10 μl mixture containing 0.2 μl of SRAP PCR products that were added with a 384-pin replicator and other components were the same as those in SRAP PCR. Tagging PCR was run with a specific PCR program: 94°C for 3 min, followed by 30 cycles of 94°C for 1.0 min, 50°C for 1.0 min, 72°C for 1.0 min and final extension 72°C for 10 min.
All SRAP primers and tagging primers were list in Additional file 2.
Pooled PCR products for Illumina's Solexa sequencing
All tagged PCR products in all 202 384-well plates were pooled and one round of size selection was performed. The PCR products with approximately 100 to 400 bp size were collected from an agarose gel and purified with Qiagen spin filter columns.
Illumina's Solexa sequencing of the pooled sample was performed by BGI (Shenzhen, China) and 13.8 million paired-end sequences (1.25 GB sequence) were obtained [GenBank:SRA035245].
Data analysis of Solexa paired-end sequences
To count how many Solexa paired-end sequences in a sequence file (text file), the sequence file was loaded in the Ubuntu Linux operation system. The 'word count' (wc) function was used to count the number of lines contained in the sequence file. Since every four lines in the sequence file contained one paired-end sequence, the total number of paired-end sequences was calculated.
Brute-Force Algorithm was used in a Java program to find the primer sequences in each paired-end sequence. With a maximum of one mismatch allowed, the primers were identified in the first 25 nucleotides of the sequences. The identified primer sequences were then removed from the sequences and the primer names were assigned to the paired-end sequences. Since all the sequences were paired-end ones, only those in which both forward and reverse primers were found were included in the following analysis.
The sequences were sorted and assigned to different files based on the forward primers. Therefore, the large sequence files were divided into 216 files that corresponded to all tagging primers. These 216 files were classified into three groups that corresponded to three sets of SRAP primers. These three groups were analyzed respectively. Java programs were developed to count the duplicates of every unique Solexa paired-end sequence in each file.
Construction of an ultradense genetic map
To integrate Solexa sequences on a genetic map, SRAP technology was used to construct an ultradense genetic map. In total, 805 SRAP primer pairs were used to run the mapping population. Six hundred and eighty-two out of 3,072 pairs from the first and third sets of SRAP primers used for preparing Illumina's Solexa sequencing sample were selected. Two hundred and forty-one SRAP primer pairs corresponded to those used in the construction of a previous genetic map of B. napus and 175 out of 241 SRAP primers overlapped with the above 682 SRAP primer pairs. Additionally, fifty-five SRAP primer pairs from forward primer 'ME2' (5'-TGAGTCCAAACCGGAGC-3') and other forward primers were combined with reverse primers of the first and third sets of SRAP primers.
The same RI line population in Illumina's Solexa sequencing was used for mapping. All forward primers were fluorescently labelled. The set-up of SRAP PCR reactions and running program were the same as those used in the first round of SRAP PCR for preparing Illumina's Solexa sequencing samples. The PCR products were separated with an ABI 3100 Genetic Analyzer (Applied Biosystems, California, USA). The data were first analyzed with ABI's GenScan software and then loaded into Genographer software for scoring polymorphic loci.
JoinMap version 3.0 Software was employed to construct a genetic map. Since there were thousands of SRAP markers, all markers were first classified into 10 groups at a high LOD score. Second, each group was divided into parts of which each has less than 200 markers. Third, a third round map for each part was generated and closely linked markers with a genetic distance of 0 to 2 cM (depending on the marker number of a group) in each part were removed. Then, the remaining markers from all the parts of a group were put together to generate a linkage map for the group. Closely linked markers with genetic distance of 0 to 2 cM were removed again. All markers with a genetic distance of 0 to 2 cM were put into a bin. Finally, when those previously removed markers were added to their corresponding bins, a genetic bin map was assembled.
Matching Solexa sequences with SRAP markers
A SRAP profile produced by a pair of SRAP primers always included multiple fragments with different sizes. Each fragment theoretically corresponded to a specific locus in the genome. A Solexa paired-end sequence was located at the two ends of a SRAP fragment. To match Solexa sequences with their corresponding SRAP markers on the ultradense genetic map, locus-specific primers were designed using Solexa sequences (Additional file 3) [GenBank:SRA035245]. The ends of these Solexa paired-end sequences containing reverse primers were used in the primer designing. After the primers were removed from these Solexa sequence ends, the left sequence parts corresponded to specific loci and one sequence, to one locus in most cases in the genome were used for the locus-specific primers. These locus-specific primers were combined with the original forward primers to amplify the SRAP products obtained from sixteen RI lines and the original SRAP forward and reverse primers in the mapping procedure. The PCR set-up and running program was the same as the tagging PCR described previously. These labelled locus-specific PCR products were separated with the ABI 3100 genetic analyzer. The fragments and their sizes were scored and compared with the SRAP markers that were generated with the corresponding SRAP primers.
Direct integration of Solexa sequence data on the genetic map
The second and third set of SRAP primers were used to amplify 94 DNA samples for the whole mapping population and SRAP PCR products were tagged with two sets of tagging primers to distinguish the DNA samples. Therefore, it was possible to integrate Solexa sequences onto the SRAP genetic map on the basis of Solexa sequence frequencies in each DNA sample. First, all tagging primer and SRAP primer sequences were eliminated from the Solexa paired-end sequences to retain the sequences that corresponded to the sequenced genomic parts. Then, each unique trimmed paired-end sequence was searched and counted in the previously established ninety-four files that corresponded to the ninety-two RI lines and the two parents of the mapping population. Finally, all unique trimmed Solexa paired-end sequences corresponding to individual loci in the genome and the absolute numbers of each trimmed unique Solexa paired-end sequences represented continuous variation of PCR amplification at a locus.
Statistical significance tests of linkage analysis between each unique Solexa paired-end sequence and 465 markers of the genetic bin map were carried out by Windows QTL Cartographer software 2.5 (http://statgen.ncsu.edu/qtlcart/WQTLCart.htm) using 'single marker analysis' function. The threshold was set at LOD score 2.5 (empiric QTL confidence intervals 99%). The unique Solexa sequences having significant linkage were assigned to the bin that had the maximum LOD score. Unique Solexa sequences which had no significant linkage to all bins or significant linkage to several bins were discarded. Thus, unique Solexa paired-end sequences were assigned into their corresponding bins on the genetic map after statistical significance testing. Compared with SRAP markers, some integrated Solexa sequences were identified as closely linked to SRAP markers that were produced with the same SRAP primers.
Alignment of the current genetic map with other published genetic maps
There were 243 common SRAP primer pairs that were used to construct both the current genetic map in B. rapa and the previously published genetic map in B. napus. Since 10 B. rapa chromosomes have their corresponding counterparts in B. napus, the same SRAP markers may exist in both genetic maps. All SRAP markers that were produced by the same SRAP primer pairs and which had similar fragment sizes were identified to align the current B. rapa genetic map with the previous B. napus genetic map. Additional SSR markers used previously were used to align the current map with other published ones.