Human contaminants in viral metagenome

[fefe89]

Hello everybody

I'm analyzing three metagenomic samples from viruses extracted from river waters.

Before sequencing samples underwent tangential flow filtration step/ CsCl ultracentrifugation, Dnase treatment and check with qPRC.

In only one of the three samples I found a massive presence of human reads (around 80%, the other 20% were viral and bacterial reads), while very little presence was observed in the other two samples (< 2%).

The BLAST analysis of these reads usually give >95% identity against human.

Can anyone suggest me an explaination?

Thank you to everybody

[Brian Bushnell]

Human contamination of environmental samples is usually due to improper handling when sampling or in the lab. It's common, particularly when high levels of amplification are performed due to low input DNA. Though if you sampled from a river where dead bodies are floating around I suppose the explanation might be different...

[fefe89]

Uhm..a dead body... I actuallay found one in my sample :P

By the way, the thing I don't understand is that even if the three samples were processed and sequenced always togheter, only one showed very high contaminants level, while the other almost nothing.

Do you think it is possible?

[pmiguel]

What was your "check with qPCR"?
It could be that high-human sample started off with 1000X more human DNA in it that the other 2 samples. Each purification step would have removed some of the human DNA, but not all.

What type of sequencer/chemistry was used?

--
Phillip