Exactly what I think. The problem is: They think they know it all. They don't imagine that maybe there could be something that they didn't think about, or something that they were wrong about, which could make a difference. Or that perhaps it's more than one thing (some of the studies, for example, DID find XMRV, just in very very low percentages - So, perhaps, had they have a more strict definition for the patients, they would have found it more - and had they combined it with the best assays, they would have found it in percentages that are similar to those found by the WPI/NCI and NIH/FDA). Bob, I think I disagree with you about "we better not have full replication studies". I think that the problem is with scientists who say that "replicated the WPI/NCI's work and didn't find XMRV", when in fact they haven't replicated the methods. Full replication means full replication. I guess that at least at first we shell not require 100% full replication, and just 99% full replication - and that is because there is at least one thing that is not reasonable to do right now: To take patients' samples and store them for 2-3 years (as the WPI used for their "Science" study samples that were collected between 2006-2008). We should, however, demand, probably, that they would wait at least a couple of days between collection and testing - as it was shown in the BWG's work, until now, that it might make a lot of difference. But in all other aspects - they can replicate. Perhaps we cannot demand them to by the same PCR machine (I mean - the same type from the same brand), but we can ask them to include the same kind of patients, collect the blood into the same kind of tubes, process it as it was done in the "Science" paper, test it as it as done in the "Science" paper and with all of the tests used there, and also use tools and kits which are the from the exact model and brand that the WPI/NCI used in the "Science" paper. And also: If they think that they don't understand something, or don't know how to do something - or that there is a slight chance for that - they must contact the WPI/NCI for guidance. If they have done that and found nothing - well, I would still think that they should collect samples, store them for a couple of years, and test them again - with all of the methods used in the "Science" paper (and with same type of patients and stuff like that - everything the same, this time, 100% replication, perhaps besides the model and brand of the PCR machine) - but the chance of that giving a different result would be small, perhaps tiny. But so far, no one has tried to fully replicate the WPI/NCI study.