I am assembling a transcriptome obtained from the Internet. The transcriptome was extracted from a human cancer tissue that had been previously grafted into a mouse. I have detected that many assembled RNA sequences do not match precisely to their reference but have additional sequences at one of their tails containing bacterial DNA. It's not that there are reads pertaining to that bacterium (always the same one) but that the RNA sequences of the cell have incorporated fragments of bacterial sequences (the reads corroborate these insertions with high coverage), therefore I can't just filter out bacterial reads as there are not reads exclusively pertaining to bacteria but hybrids. My question is how can it be possible? What possible mechanisms could have brought these bacterial fragments to the RNA of the cell. I have read in some papers that there is speculation that bacteria could be involved in the genesis of cancer cells. That would be an explanation if it wasn't because there are also mouse RNA for which the same circumstance applies: they also have bacterial sequences at the tails. I have also read in papers that there is speculation with the possibility that cancer cells are more susceptible to get DNA from bacteria transferred to their DNA. I am speculating that is the reason in this case (and somehow that weakness was transferred to contiguous mouse cells) but I would like to know if there is another possibility I am missing. Thanks.
From time to time these situations are encountered, notably when PCR is being deployed in rare DNA samples.
In this instance the wiki states,
Ralstonia solanacearum is an aerobic non-spore-forming, Gram-negative, plant pathogenic bacterium.".
I'd leave it. The simple explanation is there has been a bad mix up. The best explanation is the people loading the Illumina machine, designating the adaptors etc ... In this scenario it has nothing to do with bacterial genome integration, plant bacteria don't touch mice and aerobic bacteria wouldn't grow in cancer cells.
The second issue is what is a soil bacteria doing in a mouse cancer biological facility? For the staff this wouldn't be good news.
The singular biological possibility is its a Pseudomonas and some species are serious opportunistic pathogens of immune-weakened humans, but there are a lot of Pseudomonas species and this one doesn't ring any bells, BUT its mice so who knows? One species in particular is associated with hospital outbreaks, but a biomedical unit that isn't quite the same thing.
The pathogen in humans is called Pseudomonas aeruginosa. If hospitals get this infection it is BAD NEWS, for example burns victims are really susceptible and without intervention it kills them. If you have evidence that the bacterial equivalent in mice is the bacterial pathogen you have found that is a substantial step. Biological Safety Facilities (BSFs) DEFINITELY get infected. They hate it when this happens and can be very picky for fear of a contamination, but it certainly happens. If you have evidence of a major contaminent pathogen in the long run it would be of HUGE help to BSFs globally. No-one really bothers with the disease pathology of mice - because their mice and spend their entire lives going 'squeek' - but when they are models for cancer their status changes beyond recognition. So I'd erase my initial thoughts and suggest, what you have found is publishable and the application is BSF contamination - don't say that though say this could be the mouse equivalent of P. aeruginosa. Its certainly a publication and goodluck.
What exactly is the sequence? Are you sure it's not an artifact of the library prep?
5 seconds of googling turns up this:
is on the adapters.
So is this
Bacterial contamination would not put bacterial sequence at the end of host transcripts! But adapter contamination sure could be.
Here's the index 2 adapter
Here's your sequence
Here's it's blast against nr
It's all adapter.
Whoever did the transcriptome assembly didn't trim for adapters first. That's all.