To study the impact of wheat streak mosaic virus on global gene expression in wheat curl mite, we generated a de novo transcriptome assembly using 50 x 50 paired end reads from the Illumina HiSeq 2500. Reads were assembled using Trinity (version 2.0.6) and contigs greater than 200 nt were retained. All assembled transcripts were annotated using the Trinotate pipeline using blastp searches against the Swiss-prot/Uni-Prot database, blastx searches against the Swiss-prot/Uni-Prot databases, HMM searches against the Pfam-A database, blastp searches against the non-redundant protein database, and signalP and tmHMM predictions. To reduce noise from low abundance transcripts not well supported by the data, we filtered the assembly to retain only those transcripts with TPM values >=0.5.