Hacker News Reader

Comment by fifnir | Hacker News Reader

> activation of retro viruses during embryonic development.

Citation on this? As far as I know (am writing a phd thesis on development++), there's nothing like "virus activation". Viral elements (mostly transposons) are big evolutionary players but are usually actively repressed by DNA methylation (in eucaryots).

> Which might be why many cancers, seem, to me, to look like distinct entities, like the tissue has suddenly switched to a different functional mode.

I don't 100% understand your comments about viruses here, but related to this, we don't need to imagine some virus orchestrating such a thing (besides, as far as I know they are way too simple to drive complex eucaryot gene regulation), but rather gene regulatory networks which if they get unbalanced can lead to significant changes in morphology.

<edit> some spelling

View on HN

"The early embryo is a hotbed of activity for endogenous retroviruses, recent studies have shown. To understand why embryonic cells make viral proteins, scientists have run experiments to see what happens when viral genes are silenced.

These experiments suggest that viral proteins help the embryo develop a variety of tissues. "

From the op article.

I'm am not an expert in this at all, these are just my thoughts on the matter, take them or leave them.

Regarding regulatory networks binding together elements so as to orchestrate changes in morphology - I had not thought about this, of course you are correct. Where you say that the viruses are too simple to drive sufficiently complex changes to the regulatory networks, I would counter by saying that chance mutations are sufficient to drive these changes.

Link?

"From the op article"

We have a LOT of "appropriated" viral code in our own DNA: "Reverse-transcribed RNA molecules compose a significant portion of the human genome. Many of these RNA molecules were retrovirus genomes either infecting germline cells or having done so in a previous generation but retaining transcriptional activity. This mechanism itself accounts for a quarter of the genomic sequence information of mammals for which there is data."[1]

But that they are now part of us and when they are activated it's not some virus waking up but parts of viruses that we have adapted into doing useful things for out cells.

Viruses can cause significant changes in gene networks (for example oncoviruses usually suppress tumor-suppressing proteins [2] which dramatically affects things downstream), but the changes are not as holistic as I think you are describing them.

Yes point mutations (and viruses) can derail neural networks and that can have far-reaching consequences, but I think these things are accidental in both cases and not some virus plan as I felt you were arguing.

[1] https://www.sciencedirect.com/science/article/pii/S136952741... [2] https://en.wikipedia.org/wiki/Oncovirus

You are probably right. You know far more than I, but I do have a knack for seeing paradigm shifts coming. Take that as you will, I can't prove it, it's just personal experience. Just try to remember, if this all unravels in favor of my position, that you talked to some guy on the internet who saw it coming. It's getting pretty annoying having called it ahead of time, and no one remembering or caring that I said anything.

I'll bite: could you cite some instances of real paradigm shifts that you have previously predicted in public?

Reply on news.ycombinator.com