Warning: this post contains speculation going beyond even my usual ideas. Some of it is based on current research, but other parts are wild ideas. While you can find "quotes" from viral sequences all over human DNA, the non-coding regions sometimes dismissed as "junk" are especially notorious for holding viral sequences. Some of these are from retroviruses, but by no means all. How could sequences from RNA viruses enter the DNA genome? Those viruses must have been present in hosts also infected with retroviruses. New work has shown that far more of the "junk" is active in controlling expression of coding genes than previously thought. This should have been apparent earlier. For example, in terms of coding DNA there is little difference between humans and chimpanzees, and many of those differences are simply random genetic drift. The big differences are in epigenetic control of genes. This determines when and where they are active. At this point I feel the need to reiterate something about viruses: they don't do much of anything by themselves. Virions can often be crystallized and stored. They have no metabolism. Virtually every action attributed to "the virus" is actually performed by host cell machinery. Viruses are the ultimate puppeteers. Some ERV elements are active in healthy mammals during pregnancy. A very plausible scenario for the origin of placental mammals starts with a persistent viral infection which altered the production of complete eggs. Similar parallel evolution of live-bearing has apparently taken place in a number of separate animal clades with no recent common ancestor. (It is even possible viruses played a role in what I regard as the big historical mystery of biology on Earth, the origin of eukaryotic cells.) I'm going to go out on a limb, and speculate that viruses in general, and retroviruses in particular, are primarily responsible for the extensive development of epigenetic control seen in cells of all metazoans. We tend to blame viruses for taking over control of our cells. I'm proposing they were involved from the beginning. Our distant ancestor cells simply co-opted mechanisms first discovered by viruses to use for their own purposes. This isn't as paradoxical as it may seem, because in-the-long-run survival of the virus depends on survival of the host species, if not the individual host. If true this would make many convenient distinctions, like endogenous and exogenous, inheritance and environment, or (normal) developmental and pathological, much less clear. This will upset many, but it reinforces something we should all keep in mind: biology does not respect human categories. Nature simply does; it is humans who categorize.