Revealing genomic variation in perennial ryegrass (Lolium perenne)

Perennial ryegrass (Lolium perenne) is one of the most widely cultivated grass species and is an important crop for forage production and pastures. Understanding the relationship between phenotypic and genomic variation can help to improve breeding strategies. This requires a catalogue of gene variants based on computationally predicted effects of genomic polymorphisms. The L. perenne genome displays high levels of heterozygosity due to the outbreeding nature of this species. We used a targeted resequencing by probe capture approach in a collection of 746 perennial ryegrass accessions for 500 candidate genes that are known to regulate important biological processes such as plant growth, development and architecture. Here, we present a strategy to identify the genomic variation when dealing with highly divergent sequences. De novo assembly of the short reads led to the independent reconstruction of many divergent alleles per gene in the breeding genepool and across natural accessions, revealing highly relevant alleles in particular gene families. Traditional mapping and variant calling approaches are blind for these alleles, because the respective short reads are too divergent to be mapped onto the reference genome sequence. Using an intermediate step of de novo assembly made it possible to establish a reliable catalog of genomic variation for 500 candidate genes and their closely related paralogs. This data will be aggregated into a genomics resource containing the natural genomic variation within each candidate gene that will be used to establish genetic associations between phenotypic traits and genetic variation. Moreover, this strategy can be useful to reveal the genomic variation in other outbreeding crops to cope with high levels of heterozygosity, or for specific highly divergent gene families.