Teosinte unigenes' sequences

Due to the lack of available reference genome, clean reads from Zea may ssp. parviglumis and Zea mays ssp. mexicana were first de novo assembled into contigs using Trans-ABySS, which can use a multiple K-mer strategy to assemble the transcriptome. After that, transcriptome from different samples were pooled to reconstruct the transcriptome using CD-HIT-EST with a word length of 8 and an identity threshold ≥ 0.98.Sequentially, CAP3 was used to discard the redundancy transcripts, and the resulting contigs and singlets were combined to represent transcriptome of teosinte.