Background: The dispensable genome of a species, consisting of the dispensable sequences present only in a subset of individuals, is believed to play important roles in phenotypic variation and genome evolution. However, construction of the dispensable genome is costly and labor-intensive at present, and so the influence of the dispensable genome in genetic and functional genomic studies has not been fully explored.
Results: We construct the dispensable genome of rice through a metagenome-like de novo assembly strategy based on low-coverage (1-3×) sequencing data of 1483 cultivated rice (Oryza sativa L.) accessions. Thousands of protein-coding genes are successfully assembled, including most of the known agronomically important genes absent from the Nipponbare rice reference genome. We develop an integration approach based on alignment and linkage disequilibrium, which is able to assign genomic positions relative to the reference genome for more than 78.2 % of the dispensable sequences. We carry out association mapping studies for rice grain width and 840 metabolic traits using 0.46 million polymorphisms between the dispensable sequences of different rice accessions. About 23.5 % of metabolic traits have more significant association signals with polymorphisms from dispensable sequences than with SNPs from the reference genome, and 41.6 % of trait-associated SNPs have concordant genomic locations with associated dispensable sequences.
Conclusions: Our results suggest the feasibility of building a species' dispensable genome using low-coverage population sequencing data. The constructed sequences will be helpful for understanding the rice dispensable genome and are complementary to the reference genome for identifying candidate genes associated with phenotypic variation.