The Global Landscape and Functional Impact Of Copy Number Variation In Maize.

The highly variable maize genome is saturated with copy number variations (CNVs) which may underlie tremendous phenotypic variation. A reduced representation sequencing approach was applied to genotype 14,135 maize inbred lines. By combining association mapping approaches with machine learning algorithms, 1.4 million small CNVs were characterized. These CNVs were more prevalent in pericentromeric regions and were positively correlated with repeat density and negatively correlated with recombination rate and gene density. Large CNVs were detected by coverage, accounting for 89% of the genome. Genomic regions with high amounts of CNVs were enriched for genes within stress and stimulus responses, depleted for those in biosynthetic and metabolic processes. Conducting a genome wide association study (GWAS), we found that CNVs were enriched for associations with traits.  We postulate that CNVs associated with regulatory elements are a significant source of adaptive variation in maize.