I've been sequencing, assembling, and annotating quite a few liver and spleen transcriptomes from wild rodents, and I have noticed that many of the assembled transcripts don't get blast hits to other rodents or mammals. Through some broader blasting, I've discovered that some of the transcripts may be viral. Can anyone recommend a good database (that I can download) to blast against to try to specifically identify viral transcripts? I'm hoping for a database that excludes everything except potential pathogens.