Question about VIRTUS2 Output

[ppark123]

Hello, I ran VIRTUS2 with single end RNA-seqs from this GEO accession: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE155925. The resulting txt file isn't showing the virus that the sample is infected with, and instead shows other viruses. It seems like Hepatitis C virus genotype 1 is found in almost all the samples that I ran VIRTUS2 on, although none of the samples were infected with this particular virus. I'm not too sure what's going on. Any help would be appreciated!

[yyoshiaki]

HCV is a typical false positive because it contains polyA in the genome. In addition to the number of mapped reads, % coverage mapped on the viral genome is informative to discriminate false positives. GSE155925 is blood samples from respiratory infection donors. Viremia is not common, so it is common that you don't observe any virus from blood samples even when the donor is infected by some viruses in some organ.

[ppark123]

Thank you for the response. As a follow-up question, are there any specific types of samples (e.g., tissues) that VIRTUS2 works best with?

[yyoshiaki]

As the positive control, the following datasets may be nice.

• PRJNA557257: flu infected lung (artificial). This is used for the example of VIRTUS_wrapper.py
• PRJNA328986: chronic hepatitis. Used in VIRTUS paper in sup note.
• PRJNA601736: COVID BALF, used in sup note. Of course, you can find other positive control datasets easily by googling.

[gailrosen]

Can you explain what virusSJ.out.tab is and what the columns are? What does SJ stand for?

[yyoshiaki]

Sorry for my late reply, the file is the output of STAR and splice junction information. This information is not used for the VIRTUS pipeline. https://github.com/alexdobin/STAR/blob/master/doc/STARmanual.pdf