Creationist Ashby Camp, writing at
TrueOrigins, quoted a scientific article referring to "insertion hotspots" that were the target of viral integration much more often than other spots in the genome. Although this is superficially a valid criticism to be hurled at the ERV argument, it quickly fails.
The very "hottest" spot Camp could dig up in the scientific literature was one that was 280 times more likely to be involved in a viral integration than we could expect from chance alone. It could be pointed out that this was a virus that
doesn't exist in nature, but was specifically designed to facilitate gene therapy by targeting a specific part of the genome and replacing a crippled gene with a functional copy.
But let's ignore that for now, and assume Camp did
not misrepresent relevant research. Just how large is this in the big picture? The human genome is 3 billion or so nucleotides long (
Source). The chance, then, of randomly inserting into the same section of the genome is
1 in 3 billion.
Now let's assume there are hot spots an unprecedented 1,000 times
more likely to be attacked by a virus than the unique, genetically engineered one Mr. Camp was able to dig up. Divide 3 billion by 280,000, and you find the chance of an independent insertion is one in around 10714; this is an unlikely enough figure to be negligible, and it was derived from estimates
orders of magnitude more liberal than the data would allow!
Additionally, this doesn't address
any of the other factors involved. It doesn't explain why the
same viral fingerprint would be left behind, how it would affect the one germ line cell out of millions that advances to zygote status, how that single individual would enjoy reproductive success sufficient to establish his ERVs in the population, etc.
For the reasons listed above, it's astronomically improbable.
Not only would the same exact retroviral insertion have to occur independently at the same exact locus of a uniquely lucky sperm or egg cell that survived to adulthood and achieved enough success to establish the insertion in their respective populations, but this impossible set of coincidences would have to occur for 3, 4, even a dozen species at the same time, and for every single ERV shared between them to boot!