I don't think that for contamination to be a problem you would expect 100% positives. The studies that did find they had a problem with contamination weren't getting 100% positives. If you just need one cell of mouse DNA to get a false positive, there's going to be a lot of room for randomness affecting results.
Wasn't that because they examined different isolates? In only two of them they found laboratory contaminants. I'd expect if you checked those two isolates 100 hundred times you'd find find 100 times laboratory contaminants.
Why wouldn't this hold for the Mikovits/Alter/commercial labs testing? I know it's speculated that the former two used differences in handling of the samples between patients and controls. While I really doubt this, this absolutely can't be true for the commercial lab testing because they have no idea who's a patient and who's not.
re:
why mouse contamination wouldn't be a problem for EBV etc studies: Couldn't it just be that mouse DNA doesn't contain sequences which could be interpreted as EBV, but they do for XMRV?
I understand, and XMRV is a mouse virus afterall. But since this virologist said mouse DNA contamination is so common they have to get that from somewhere, don't they? Where are those studies? (this is a rethorical question because I know they exist, but are they representative? In other words, did they use studies that examined the methodology/tubes used in common practice, like testing for EBV/HIV etc.)
ps: I'm just guessing with this, but find trying to answer other's questions is a good way of testing my own knowledge. Please feel free to correct me if I'm wrong.
Offcourse, since most of us gain the knowledge because of this terrible disease and not in a professional setting it's quite hard to understand all of this. These discussions make us all smarter, and heck we even outperform most virologist by our common knowledge. (for example Vincent Racaniello retracted his statement after this forum discussed these studies)