High-resolution genetic mapping of maize pan-genome sequence anchors
Fei Lu, Maria C. Romay, Jeffrey C. Glaubitz, Peter J. Bradbury, Robert J. Elshire, Tianyu Wang, Yu Li, Yongxiang Li, Kassa Semagn, Xuecai Zhang, Alvaro G. Hernandez, Mark A. Mikel, Ilya Soifer, Omer Barad& Edward S.
Buckler;NatureCommunications6,Articlenumber:6914,doi:10.1038/ncomms7914,Received10 February 2015 ,Accepted11 March 2015 ,Published16 April 2015
Genome duplication1 and transposable elements2 (TEs) are important driving forces behind plant genome evolution, and have generated the complex genomes found in many major crop species3, 4, 5, 6, 7. These complex genomes contain tremendous structural variation (SV), in the form of copy number variation (CNV), presence/absence variation (PAV, an extreme form of CNV), inversion and translocation. In humans, CNV has a limited influence on disease susceptibility and explains only a minority of the ‘missing heritability’8, 9. However, in major crop species, CNV is much more prevalent10, 11, and thus is much more likely to significantly have an impact on phenotypic variation. For example, plant disease defense genes often display CNV and frequently colocalize with other CNVs12, 13. Furthermore, read depth variation is over-represented in genome-wide association study (GWAS) hits for multiple traits in maize14. These observations suggest that CNV plays an important role in phenotypic variation.
To characterize CNVs, an ideal system is the pan-genome, a representation of both the core genome (collinear genome) and the variably distributed genome (SVs) of a species15. The pan-genome can be constructed by comparing multiple genomes derived from de novo sequence assembly (Fig. 1). As a result of the falling cost of sequencing, it is now possible to
sequence crop varieties on an unprecedented scale in both depth and sample size. However, because of the repetitive nature of complex genomes, prevalent alignment ambiguity hinders accurate read conting and confounds pan-genome assembly. In addition, a large proportion of genomic fragments absent from the reference cannot be placed on the pan-genome via alignment. Therefore, for species with complex genomes, sequence alignment alone is insufficient to build high-quality pan-genomes. However, the availability of a set of ultrahigh-density genetic anchors would be extremely helpful to the pan-genome construction. These genetic anchors could be used either to evaluate assembly quality or, even better, to direct de novo assembly of individual genomes6. Genotyping-by-sequencing (GBS)16, a reduced representation approach, can efficiently generate abundant single-nucleotide polymorphisms (SNPs) for a large number of individuals of a species. It is also a cost-effective source of sequence tags that can be used as genetic anchors to direct contig/scaffold assembly and to map genomic fragments absent in the reference. In this study, we developed an efficient and accurate approach to genetically map ultrahigh-density sequence anchors, which will be a valuable tool for ongoing pan-genome construction. This approach is most powerful with the large sample size of individuals afforded by GBS. This analysis was conducted in maize, the largest production crop in the world, which is also a model species for complex genomes. Mapping sequence anchors in maize provides an effective example for other species.
