De novo genome assembly of reveals rapid genome expansion and adaptive evolution.

The wild relatives of rice have adapted to different ecological environments and constitute a useful reservoir of agronomic traits for genetic improvement. Here we present the ~777 Mb de novo assembled genome sequence of . Recent bursts of long-terminal repeat retrotransposons, especially , led to a rapid twofold increase in genome size after speciation. Universal centromeric tandem repeats are absent within its centromeres, while -type LTRs constitute the main centromere-specific repetitive elements. A total of 40,116 protein-coding genes were predicted in , which is close to that of . Both the copy number and function of genes involved in photosynthesis and energy production have undergone positive selection during the evolution of , which might have facilitated its adaptation to the low light habitats. Together, our findings reveal the rapid genome expansion, distinctive centromere organization, and adaptive evolution of .