[biotechbio]

Sure, a 2020 study from the Pan-Cancer Analysis of Whole Genomes (PCAWG) Consortium, looking for viral evidence in thousands of tumor genomes and transcriptomes[0]. Part of a massive, cross-institutional effort.

"Searching large pan-cancer genome and whole-transcriptome datasets enabled the identification of a high percentage of virus-associated cases (16%)".

Far from majority.

[0] https://www.nature.com/articles/s41588-019-0558-9

[panabee]

thanks for sharing. will dig into the methods.

based on ebv studies i have read (happy to share if you want), some papers use flawed methodologies for viral detection (e.g., checking for limited set of viral proteins).

to reiterate, we mostly agree, except i adopt a more restrained stance: the conclusion supported by science is that viral causation is provable in some cancers -- but not a majority.

which is a subtle, but crucial difference, from concluding that viruses do not cause a majority of cancers (much higher bar IMO).

for instance, past studies may have used flawed detection methods or extrapolated from unrepresentative samples like the lung cancer study shared earlier.

[biotechbio]

Sure, absence of evidence is not evidence of absence. But these PCAWG results have been discussed to death since they were published, and its pretty sound science.

You could also take the bottom-up approach of asking what DOES cause certain cancers. That's a whole other discussion.

Considering all this, if you still have doubts that "viruses do not cause the majority of cancers", I think you will likely be skeptical about pretty much all of biology.