Hi, I have used Salmon on some new RNAseq data that includes a host and virus (human and virus). When indexing the transcriptome, I concatenated the virus genome (as they don't have transcriptomes) to the .fatsa file of the host. I did the same with the .gtf file used for building the txdb object. However, following the tximport the virus is not included in the counts object. I believe the issue is when constructing the tx2gene tibble, as there is no TXNAME for the virus...

#construct a tx v. gene name tibble

k <- keys(txdb, keytype = "TXNAME")

tx2gene <- AnnotationDbi::select(txdb, k, "GENEID", "TXNAME")

Any suggestions for a work around for this? I would really like to compare the virus counts per sample :)

Thank You!

Let me just explain what tximport does for gene summarization: for every transcript in the quant.sf file, it looks up the gene name, and then collapses the information from the transcript level to the gene level. If you want the viral transcripts to pass through without being summarized, you should add rows to the tx2gene table which have the transcript name in the first and second column.