The Teosinte (Zea mays ssp. parviglumis) de novo Genome Assembly

Maize (Zea mays ssp. mays) was domesticated from the wild grass teosinte (Zea mays ssp. parviglumis) approximately 9,000 years ago in the Balsas River Basin of southwest Mexico (Matsuoka et al., 2002). Insights regarding gene loss, gene gain, and large-scale structural evolution during domestication can be discovered by comparing the genomes of teosinte and maize. Characterization of these phenomena will improve our understanding of the domesticated phenotypes plant breeders fine-tune to improve maize production. However, to effectively compare maize and teosinte, high-quality teosinte genome assemblies must first be generated. Here we report our assembly efforts to produce draft genomes for two teosinte inbred lines (TIL01 and TIL11). We used three different de novo assemblers (RAY, ALLPATHS-LG and MaSuRCA) on the XSEDE High Performance Computing clusters to generate the draft assembly. Our preliminary results indicate that MaSuRCA has performed better in terms of total genome assembly length, CEGMA genes coverage and other assembly metrics. The assembled scaffolds for TIL01 (1.2 gigabases ) and TIL11 (1.8 gigabases) cover approximately 50% and 75% of the teosinte genome. Estimated repetitive content of these genomes ranges from 80 to 85 per cent. We plan to use additional long reads (PacBio) and RNA-seq data to improve this assembly.