Viral load estimation from WXS (whole exome sequencing) data

Question

[akhattri]

Is there any method of estimating number of viral genomes per host genome using whole exome sequencing data. I am specifically interested in HPV which integrates to host genome in head and neck cancer and also in cervical cancer.

Thanks

[WouterDeCoster]

Hello akhattri!

It appears that your post has been cross-posted to another site: http://seqanswers.com/forums/showthread.php?t=73161

This is typically not recommended as it runs the risk of annoying people in both communities.

[akhattri]

Yes I did post it there first but as no one replied I ended up posting here. I am trying to delete that post but unable to find option.

Answer 1

[Brian Bushnell]

It sounds like the most straightforward approach would be to map the reads to the genome and viral genomes simultaneously, then calculate the coverage ratio between host and virus. However, there's no reason to expect viral DNA in exome sequencing because the baits wouldn't be designed for it, so I think you'd need whole-genome sequencing. Exome-capture messes up abundance calculations anyway due to differential bait efficiency, so even if you designed baits for your viruses, you'd need calibration.

[akhattri]

Yes that was my concern also. The thing is data is publicly available (TCGA) so I can not do anything. I was thinking if there is anyway to still do it with exome data.

[Brian Bushnell]

You can still try the approach I mentioned; some fraction of the viruses may integrate into exons, and the part of the exon at either end of the viral insertion may still bind to the baits, retaining some of the viral genome. Exon-capture routinely captures stuff outside of the baits - there's not a coverage cliff at the edge of the baits, but a gradual decline. Also, exon-capture is not 100% efficient so you still get some coverage of the entire genome. Normally it's very low but that depends on the library prep.

Without calibration it would be very hard to get accurate estimates from this approach, though. There's lots of public WGS data out there.

[WouterDeCoster]

Please use ADD COMMENT to answer to earlier replies, as such this thread remains logically structured and easy to follow.

[WouterDeCoster]

I reverted your deletion of this post, since that would also delete the answer from Brian on this post.

[akhattri]

Yes that was my concern also. The thing is data is publicly available (TCGA) so I can not do anything. I was thinking if there is anyway to still do it with exome data.

Answer 2

[cristina_sabiers]

I don't know if this help, but maybe you want to take a look

http://www.actrec.gov.in/pi-webpages/AmitDutt/HPVdetector/hpvdetectoruserguid.html

[akhattri]

Seems it would only detect presence of HPV. I wanted to estimate the viral load.

[WouterDeCoster]

You wrote:

I am specifically interested in HPV

But detecting and definitely quantifying virus in exome sequencing is a rather bizarre strategy. As written by Brian, by design exome sequencing will not include viral sequences, so only by accident you could find some.