Marker-assisted breeding strategies, feasible in diploids inbreeding crop species, cannot be applied in outbreeding heterozygous diploid and polyploid crops because the existing genotyping technology does not provide enough resolution for pursuing the associated marker. The major reason is that currently markers can only be associated with traits at the SNP level and not at the level of an individual haplotype. This makes it impossible to determine whether a trait-linked SNP is on the beneficial haplotype, or also on another negative contributing haplotype linked by chance due to the structure of the population.
In the latter case, further breeding often results in a loss of predictive power for the trait-linked SNP, and failure of breeding for the aimed target. As a consequence, only a limited number (8-10) of selection markers are routinely used, only for qualitative traits where one allele is sufficient for expression of a dominant trait such as disease resistance, while no markers for quantitative traits, such as yield and quality traits, are used. Although the obvious solution is breeding on the level of individual haplotypes, the technology to do so is currently lacking.
A promising technology to deal with haplotype assembly in a cost-effective manner is 10x Genomics. At the beginning of 2016, the company launched their Chromium platform. The core technology is a Gel bead inside Emulsion (GEM) and uses droplet microfluidic technology prior to standard (Illumina) sequencing to generate linked reads that can be used to piece together haplotypes. Examples of success currently focus on the diploid human genome (www.10xgenomics.com). We propose to early adopt 10x Genomics technology to plant genomes and to develop methods for haplotyping and improving genome assembly customized for this technology in valuable polyploid crops.
The existing 10x Genomics bioinformatics package for data processing contains several standard public software tools (BWA, GATK) as well as specific software tools for alignment, variant calling, de novo assembly and visualization (Long Ranger, Supernova Assembler and Loupe). Adapting these pipelines to the challenges from plant species will be a priority. Right now, the 10x Genomics software is fully focused on human samples. While haplotypes can be phased to some extent for diploid (human) data, no accepted phasing method exists yet for polyploid genomes.
The implementation of accurate haplotyping in plant breeding will have a profound impact on food security and consumer product quality, by enabling haplotype-specific genome assembly for auto- and allopolyploid species, but also for heterozygous diploid species, which include major food and energy crops (e.g. potato, durum wheat, Miscanthus, yam and many ornamental crops important for the Dutch economy). This will not only address the ploidy challenge, but also decipher other unsolved issues with complex genomes: heterozygosity, structural variation, repeats and genome duplications. Building the Green Hapmap project opens new avenues for unlocking genetic diversity at the haplotype level and for facilitating optimal parent selection, to boost breeding in the Netherlands for efficient development of new global competitive varieties.