Hello, I am working with bacterial genome fasta file. I have ran Virsorter2 tool to identify the viral regions. And i have annotated the viral fasta file with Pharokka. The bacterial genome has i have are Metagenome-Assembled Genomes contigs. So the viral region identified by Virsorter2 are integrated prophages in the genome.

Now I also have scrna raw read data which I have mapped to these viral fasta and have extracted the UMI and barcode to see on a single cell level, how many viral genes or UMI map back to each cell. The scrna reads was made from a droplet based sequening method, so the UMI are barcoded hydrogel beads, so each of the forward read contains barcodes (20 bp) and unique molecular identifiers UMI (8 bp) for differentiating the single cells and gene.

The thing I am trying to see is : if CellA has multiple viral structural genes or proteins mapping like holin, capsid, terminase etc mapping to it, then it could be that these prophages are active in the bacterial genome.

Now , I think even if i find these transcripts mapping back to viral genes, the important thing to also look at to form an analysis, is for the same cell , how many transcripts are mapping to the host UMI, if for example we have 40 viral_umi vs 5000 bac_umi, then even though we see the cell having viral genes , that could mean nothing related to the host fraction.

And correct me If my logic is wrong, but how on single cell level can i know for a cell, its total_bac_umi vs total_viral umi extraction?

My workflow till now: The scrna reads are single end data that is R1+R2 info is already in the single fq file. This is how headers look like : @A00583:1030:HNMLKDSX3:1:1101:1163:1000_ACTCGCCATAGCTCTGTCTA_GACACGACThe 20 bp after here are the Barcodes and the 8Bp sequnce after that is the UMI followed by the cdna sequence in second line.

1. Bacterial genome -> VirSorter2 Output: identified viral contigs (FASTA)

2. Viral FASTA -> Pharokka Output: viral gene annotations, CDS information, GTF/GFF

3. Raw scRNA-seq reads -> BWA mapping against viral FASTA Output: mapped and unmapped BAM files

4. viral.bam + Pharokka.gtf -> featureCounts Output: read assignments to annotated viral features/genes

5. BAM/FASTQ -> UMI-tools Output: extracted cell barcodes and UMIs, grouped into unique molecules per cell

This gives us info for each single cell, how many viral genes are being transcibed and what function they have,