I think looking to the XMRV/MLV antibodies was a strength of the study, not a weakness. The more recently infected patients should have them for certain, and the older patients might as well. WPI says they find antibodies. But I think your second comment is right on target. The authors raised a point I have not seen before in print (might have been discussed at a conference though), and that is the possibility of human cell line contamination. Apparently this is known to be possible, it has happened (sorry, I definitely read/heard that somewhere but forget the source right now). No, that does not prove human infection in the patient population. Infected cell lines means that human cells grown for many generations as part of lab reagents (commercial or proprietary) have at some point been contaminated, and the contaminant continues to grow and stay in the HUMAN cells in the reagent. There are lots of human cell lines preserved and used in lab testing (we just don't hear about that much, so it sounds strange). Apparently WPI's and some of the other XMRV tests use human cell lines at some point in the assay. If some of those cell lines are contaminated, the testing for mouse DNA would be insufficient. They would have to screen every reagent used for contaminated human cells. Given that human cell line contamination is possible, this just seems prudent and I am surprised this issue has not come up before.