Finding Reads from a DNA Virus

Question

[nickeener]

Hi, I'm trying to locate viral reads from a DNA virus in Drosophila sequencing data from NCBI's SRA and was unsure whether using RNA-seq experiments would be productive or not. Obviously the virus must transcribe its DNA into RNA in order to have its viral proteins translated, but my question is: since there will be both viral RNA and viral DNA in infected cells, should I look for sequencing runs where the genomic DNA was sequenced or would using RNA-seq runs give me better results? Or should I consider both?

Thanks

Answer 1

[darren.obbard]

Both DNA and RNA will work. In our experience virus (genomic) copy number can be several hundred-fold higher than the host genome-copy number, so on experimental cultures/ infections DNA will work well. On the other hand, DNA can be prone to false positives, as large DNA viruses often contain microsatellite sequences.

RNA reads would imply active transcription, but at least some genes are expressed at high levels compared to fly mRNAs in active infections.

See https://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1007050

If you're looking for known Drosophila viruses, see http://obbard.bio.ed.ac.uk/data/Updated_Drosophila_Viruses.fas.gz, which includes several unpublished DNA viruses.

Best wishes,

Darren

Answer 2

[Istvan Albert]

since the DNA is covered relatively evenly whereas the RNA depends on the copy number of each transcript detecting from DNA ought to work more effectively. Especially if you can poinpoint the location of the insertion.

Another way to say this is that the amount of viral RNA in the sample is probably at trace levels, whereas the DNA should end up at the same coverage as all the other DNA.

[darren.obbard]

There are no integrating retroviruses (other than TEs!) outside of vertebrates. Other insertions will not be active (infectious) viruses.

[Istvan Albert]

I was not aware of that - thanks for the information