Red Rose

Rose is the most well-known and beloved ornamental plant worldwide. As in most other ornamental plant breeding programmes, molecular tools have up to now rarely been used. There is a number of obstacles to implementing molecular breeding in roses. These include its tetraploid nature, the fact that it is vegetatively propagated and that large genetic gains can still be achieved by simple crossing and selection. Moreover, a significant gap exists between research and breeding practices, impeding the application of developed genetic tools in practice. While for large agricultural crops the primary focus is on yield and other quantitative traits for both the academic community and breeders, rose research has focussed on both characteristics that are qualitative, such as presence/absence of the ‘double flower’, and some disease resistances, and on complex qualitative traits, including flower colour, scent emission, bud outgrowth1, floral development and vernalization response2. Although rose is an ideal model species for studying the molecular basis of these traits, they are easily examined by eye by breeders. For instance, scent is a complex trait that is being studied, but it is examined by an experienced nose in a breeding programme.

The worldwide rose research community is relatively small. Therefore, the development of new information and tools for rose breeding and genetics in the past few years has been an important step. The availability of rose genome sequences3,4,5 has made the identification of candidate genes for these traits possible. Once a candidate gene has been found, allelic diversity may be linked to functional diversity (see ref.6), leading to ‘candidate alleles’, which is of much use in a highly heterozygous and polyploid crop. Next-generation sequencing has also facilitated the creation of transcriptomes and large numbers of single-nucleotide polymorphism (SNP) markers, which can be assessed using the 68k WagRhSNP array for rose7 or by single SNP assays using flanking sequence information. Owing to newly developed software for dosage scoring and genetic mapping in polyploids, ultra-dense genetic maps have recently been produced in diploid5 and tetraploid rose8,9 that contain many more markers than maps from previous diploid rose populations (collated in ref.2). These maps have enabled subsequent quantitative trait locus (QTL) mapping studies10 and genome-wide association analyses11,12 in tetraploid rose, helping to identify regions of the genome that are statistically correlated with traits of interest. As the genomes and most of the resources are publicly available, rose genomics is far more accessible than ever before.

Here we discuss the opportunities for better understanding of structural variation in the rose genome and genome evolution in the genus Rosa and applications for QTL mapping, genome-wide association study (GWAS) analysis and functional analyses of traits, and for measuring genetic diversity, which, we hope, will ultimately improve the speed and precision of breeding new rose cultivars.