**2. SRAP technology**

We developed and published information on the SRAP marker system in 2001 (Li and Quiros). The original thinking was to simplify the AFLP detection procedure and increase throughput and improve reproducibility compared to RAPD. To produce a simple detection procedure, we skipped restriction enzyme digestion and ligation of target DNA fragments and adapters in the AFLP detection protocol. We designed SRAP primers in sizes similar to those in AFLP, but ran one round of PCR instead of two rounds in AFLP. To detect multiple loci with a pair of SRAP primers, we designed a special PCR running program (940 C for 1 min, 350 C for 1 min and 720 C for 1 min for the first 5 cycles and followed by 30 cycles at the raised annealing temperature of 500 C). At the beginning of PCR, the 350 C annealing temper‐ ature allowed SRAP primers to anneal to multiple loci in target DNA so that the multiple loci were amplified to produce a profile that is similar to that in AFLP. Similar to AFLP, most SRAP markers are dominant while most SNPs and SSRs are co-dominant. Compared with RAPD, SRAP used a pair of primers with 16 to 22 nucleotides instead of 10-mer short primers in RAPD, which gives SRAP a big advantage over RAPD so one SRAP primer can combine with unlimited number of other primers. Although SRAP PCR starts at 350 C an‐ nealing temperature in the first five cycles, the larger sizes of SRAP primers allowed the in‐ crease of annealing temperature to 500 C in the following cycles, which significantly improves the reproducibility in SRAP. In contrast, a low level of reproducibility in RAPD is a limitation factor. In addition, SRAP primers can be fluorescently labelled and combined with unlabeled SRAP primers so SRAP PCR products can be separated in capillary instru‐ ments such as ABI genetic analyzers.

In general, there is a difference of GC content between gene coding sequences and other se‐ quences in plant genomes. We used this difference to design two sets of SRAP primers. The forward primers contained a GGCC cassette closing the 3' end of SRAP primers that might preferentially anneal to the GC-rich regions while the reverse SRAP primer set was incorpo‐ rated with an AATT cassette that would preferentially anneal SRAP primers to introns and gene spacers so that SRAP could preferentially amplify gene-rich regions in a genome. After sequencing SRAP fragments and constructing a SRAP genetic map in *B. oleracea*, it was found that SRAP indeed amplified more sequences from genes and more SRAP markers fell into chromosome arm regions and produced fewer markers in the centromeres that were fil‐ led with AFLP markers.

formed in one round of PCR, however, a low level of reproducibility of RAPD amplification

As next generation sequencing (NGS) technologies dramatically increase capacity and throughput of DNA sequencing, whole genome sequencing of many plant species has been accomplished and most economically important crop species such as rice, maize, soybean, sorghum, potato, tomato and Chinese cabbage have been fully sequenced. Although it is still challenging to use NGS for assembling a whole complex genome such as barley and wheat, there are thousands of SNPs identified in NGS that can be used to develop molecular markers in species with complex genomes. Furthermore, NGS is directly used in SNP dis‐ covery and a few dozen genotypes can be sequenced simultaneously to assemble ultradense genetic maps. Additionally, different strategies are used to produce partial genomes that can

We developed and published information on the SRAP marker system in 2001 (Li and Quiros). The original thinking was to simplify the AFLP detection procedure and increase throughput and improve reproducibility compared to RAPD. To produce a simple detection procedure, we skipped restriction enzyme digestion and ligation of target DNA fragments and adapters in the AFLP detection protocol. We designed SRAP primers in sizes similar to those in AFLP, but ran one round of PCR instead of two rounds in AFLP. To detect multiple loci with a pair of SRAP primers, we designed a special PCR running program (940

ature allowed SRAP primers to anneal to multiple loci in target DNA so that the multiple loci were amplified to produce a profile that is similar to that in AFLP. Similar to AFLP, most SRAP markers are dominant while most SNPs and SSRs are co-dominant. Compared with RAPD, SRAP used a pair of primers with 16 to 22 nucleotides instead of 10-mer short primers in RAPD, which gives SRAP a big advantage over RAPD so one SRAP primer can combine with unlimited number of other primers. Although SRAP PCR starts at 350

nealing temperature in the first five cycles, the larger sizes of SRAP primers allowed the in‐

improves the reproducibility in SRAP. In contrast, a low level of reproducibility in RAPD is a limitation factor. In addition, SRAP primers can be fluorescently labelled and combined with unlabeled SRAP primers so SRAP PCR products can be separated in capillary instru‐

In general, there is a difference of GC content between gene coding sequences and other se‐ quences in plant genomes. We used this difference to design two sets of SRAP primers. The forward primers contained a GGCC cassette closing the 3' end of SRAP primers that might preferentially anneal to the GC-rich regions while the reverse SRAP primer set was incorpo‐ rated with an AATT cassette that would preferentially anneal SRAP primers to introns and

C for 1 min for the first 5 cycles and followed by 30 cycles at the

C in the following cycles, which significantly

C). At the beginning of PCR, the 350

C for 1

C an‐

C annealing temper‐

be used to directly sequence SNPs using next generation technologies.

limits its wide use in genomic analysis.

24 Plant Breeding from Laboratories to Fields

**2. SRAP technology**

C for 1 min and 720

raised annealing temperature of 500

crease of annealing temperature to 500

ments such as ABI genetic analyzers.

min, 350

There is wide flexibility in the design of SRAP primers. After testing the primers we used for gene cloning, we found that most of these primers worked well in SRAP amplification. In the construction of an ultra-dense genetic map in *B. napus*, we used 12 fluorescently labeled forward primers and 442 unlabeled primers to assemble a genetic map with 13,351 molecu‐ lar markers (Sun et al., 2007). Based on the SRAP markers on the genetic map, the efficiency of each labeled primer was checked by counting the average numbers of SRAP markers pro‐ duced by individual labeled primers. Some labeled SRAP primers such as FC1, BG23 and SA7 produced more than ten polymorphic loci while EM2 and DC1 produced less than six polymorphic loci. If we checked individual SRAP primer combinations, we found that there was a big difference in the numbers of polymorphic loci detected by individual SRAP pri‐ mers. Efficient SRAP primer combinations produced over 20 polymorphic loci while less ef‐ ficient SRAP primer combinations produced one to three mapped loci. Actually, we tested over a thousand SRAP primers in *B. napus* and *B. rapa* and found that most primers pro‐ duced good profiles with over 30 strong bands (unpublished data). After testing a large set of SRAP primer combinations, we selected a set of SRAP primer combinations that showed the best performance in SRAP amplification in different accessions and populations in vari‐ ous Brassica species. These SRAP primer combinations serve as a standard set and are rou‐ tinely used in genetic map construction, mapping of quantitative loci (QTL) and gene cloning in our lab. Therefore, if SRAP is frequently used in a lab, it is worthwhile to select a set of SRAP primer combinations which will enhance the effectiveness and efficiency of SRAP marker detection.

To enhance the capacity and effectiveness of SRAP technology, we combined SRAP with Il‐ lumina's Solexa sequencing to directly integrate genetic loci on the *B. rapa* genetic map based on paired-end Solexa sequences (Li et al., 2011). To achieve this objective, we used two rounds of PCR to prepare SRAP products that were pooled and sequenced with Illlumi‐ na's Solexa sequencing. The first round of PCR produced SRAP fragments using individual DNA samples from a mapping population with the same set of SRAP primer combinations while the second round was used to tail the SRAP PCR products that allowed identifying the original DNA for producing paired-end Solexa sequences. After Illumina's Solexa se‐ quencing, paired-end sequences were sorted using the sequences of tag primers and the numbers of unique paired-end sequences from each DNA samples were obtained. These numbers represent Solexa sequence frequencies in each DNA sample which was used to in‐ tegrate Solexa sequences onto the SRAP genetic map. To integrate paired-end sequences on the genetic map, we adopted a QTL mapping strategy by using Windows QTL Cartogra‐ pher software 2.5. When a paired-end sequence was found to have only one significant LOD score in a bin of the genetic map, this sequence was assigned into this bin. In total, 1737 unique paired-end sequences representing the same number of genetic loci were integrated on the genetic map. Eventually, we constructed a high density *B. rapa* genetic map consist‐ ing of 1,737 paired-end Solexa sequences, 9,177 SRAP markers and 46 SSR markers.
