Maize and Wheat Genomic Sequence Diversity Revealed using de novo Assembly and Unbiased Comparative Genomics

A quality denovo assembly with long genomic sequences from short reads has long been considered an impossible challenge for large, repetitive genomes such as maize and wheat.

The challenge has been met and overcome by the DeNovoMAGIC2 genome assembly algorithm. Using high coverage (>180X) of short Illumina reads produced from a mix of library sizes, ranging from 450 – 10,000 bp, extremely long sequence scaffolds (L90>1 Million bp) with very low (<2%) unfilled gaps were produced for maize, diploid wheat and tetraploid wheat. These assemblies covered more than 90% of the genome sequences and could be produced with exceptional speed.

The unbiased assembly of multiple maize genomes enabled us to efficiently construct a "pangenome" database. Such a database facilitates the imputation of the full genome sequence of each individual maize line, as well as its haplotype map and polymorphism diversity, from genotyping data. As such, genomic-selection, GWAS and trait mapping are more simplified and accurate.