Now it is generally accepted as Rational Wiki puts it (atheist equivalent of Answers in Genesis) “If two organisms share the same ERV, in the same location, with the same inactivation mutations, then they almost certainly share them due to common inheritance and not two separate infections.”
But I have to disagree. Their use of the phrase “not two separate infections” is deceptive, because the same infection, in two similar species during the exact time frame, would produce the exact same results that we observe, and therefore it would not be necessary that they be two “separate” infections at all (they are implying this is a claim that has been made, though I searched and could not find it...if one of you do please post a link) and would thus NOT imply an assumption of common descent.
Time out...
Let us look at your 'rebuttal'.
In response to:
"“
If two organisms share the same ERV,
in the same location,
with the same inactivation mutations, then they almost certainly share them due to common inheritance and not two separate infections.”
you write:
"...because
the same infection, in two similar species during the exact time frame, would produce the exact same results that we observe,"
When you say "same results", are you referring to the 'the same ERV, in the same location',
or 'the same inactivation mutations'?
I note that the entirety of your rebuttal was merely to state an unsupported opinion regarding a desire to 'improve' phylogenetic trees' (more on that gibberish in a moment).
Please EXPLAIN, without copy-pastes or doctored quotes, HOW, exactly, a viral genome could be inserted at the exact same locus in a chromosome of 2 different species (given that the target integration sites are literally all over the genome*), AND possess the same inactivation mutations, purely randomly.
Let us see you probability calculations, since you also write about how you engaged in all sorts of probability stuff when you 'worked in biotech'. And don't forget to document/justify the numbers you use.
In my humble opinion because researchers seek these alleged ERVs out to form or improve phylogenetic trees, lineal relationship is already a pre-supposed reality before they look (which biases the interpretation).
Humble?
OK...
So, as far as I can tell, the first paper to use ERVs to generate phylogenies was in perhaps 1995.
Papers using coding, non-coding, DNA hybridization, mtDNA, etc., had come out at least 6 years earlier, and those using amino acid sequence data even sooner (more than a decade), and the means to compute their probabilities and such have been under constant development throughout that time.
In fact, a paper came out in 1992 providing DNA-based phylogenies for the human question with exceptionally high statistical support for the trees produced (the paper has been cited 149 times), well before the first ERV paper came out. So why would there be a 'need' to 'improve' anything?
You talk about bias (that is really all you can muster, it seems) - many years ago, a 'professional' creationist showed up on a listserv that I was a member of declaring that investigator bias totally colored the outcomes of even sequence analyses.
So, I challenged him to prove it. I took a data set that I had used for an earlier paper, removed all gaps, numerically coded the taxa identifiers, and scrambled the order of the sequences in the matrix. I provided him links to free alignment and analysis programs, and asked him to do an alignment, run a phylogenetic analysis, and report his findings to the listserv and I would compare his results to my 'biased' ones.
The index card on which I wrote down the codes for the 24 taxa is still hanging on my office wall, yellowed and curling on the edges. It was about 14 years ago.
He refused to test his creationist hypothesis.
Would YOU take the challenge? I will have to update the format of the old file - do you use FASTA in your DNA analyses?
*on a related note, shortly after the big '80% of genome is functional' nonsense from ENCODE came out, I teamed up with a math prof I know to see how many binding sites of known sequence (that was essentially ENCODE's definition of function - binding sites) - and there are a lot of them - are found in a chromosome. So, I had the math prof generate an 'artificial chromosome' of 100 million BPs in length of random sequence, and I downloaded a real chromosome's sequence (about 10 million BPs in length, if I remember), and I analyzed them both for known binding sites - the results stunned me. There were MILLIONS of binding sites (including overlaps, etc.). Point is - I think the odds are going to be very much against a purely random insertions at the same locus in multiple species.