Hi omer et al,
I was able to do a tiny bit of BLASTing but only a little analysis as I am totally blasted myself! Actually all I can do is raise a few questions about BLAST sequence comparisons as presented by this website tool. As my brain is a large fried egg at the moment, I'm not sure if any of the following will be clear, or even of any use!
There is something I can't figure out at nt 152 on the Lo/Alter fragment omer posted about (accession # HQ601957.1); most of the subject sequences have a lengthy break in the alignment there (you can see it on the red graphical sequence schematics at the top of the page; it appears as a black line in the middle of the sequence). Anyone know what that means? Are they omitting any non-identities that don't align, thus inflating the percentage identity estimate? Not sure.
Leaving aside that point, there is indeed a lot of homology between this Lo/Alter gag fragment and 22Rv1, VP62, etc; nearly 100%, as omer pointed out. Keep in mind, though, that this Lo/Alter fragment is small (only 276 bp) and comes from a region of gag further downstream than the beginnings of the larger Lo/Alter gag fragments.
Also, compare the BLAST results for the larger (698 bp) Lo/Alter fragment "MLV-related virus CFS isolate CFS-type1 polyprotein gene, partial cds" (accession # HM630562.1): notice that it has less overall percent identity (96%) with XMRV isolates like 22rv1 and VP62, and seems to start after the 24 bp gag leader sequence deletion.* If you look at the same gag region from which the fragment HQ601957.1 was taken (and had close to 100% identity in sequence alignment with prostate XMRV isolates), you do find that HM630562.1 has only 95-96% percent identity with aligning sequences of VP62 and 22Rv1, respectively, which supports the idea that the smaller fragment could come from a virus that maintains high percent identity over a longer gag range...
What all this means is that there is at least one Lo/Alter gag frag that is highly similar to prostate isolates, though it covers only a small region of gag. However, at least one of their larger isolated fragments is less similar.
It would be interesting to compare the other Lo/Alter sequences to the prostate and WPI sequences as well. As G just pointed out, the env sequences would be interesting, in part because of the higher variability one might expect there under selective pressure from the immune system.
*Note: I think that the BLAST percentage identity estimates are leaving out the 24 bp deletion of "classic" XMRV isolates, since these don't align with the Lo/Alter fragments; it seems that percent identity is based on similarities in parts of the sequence that align, and does not incude those that do not. If that deletion were included, the genetic homology would of course be lower (by about 2-3% in terms of percent identity over the 698 bp gag region). I am not sure if and when deletions are considered in sequence homology calculations, however.