Well, like I said, if you approach the problem using predictive analysis. You would use patients based on "Random Forest Prediction" models that satisfied the biomarker palette for CFS patients as your subjects. You can sceen for CFS using only biomarkers, Dr. Petersen has been studying these biomarkers for decades.
I understand (to some extent anyway) about research methodologies.
But you appeared to be accusing us of not doing enough for the people who don't have XMRV.
I personally, and I imagine most people reading this thread, won't be doing any "random forest prediction" models on data any time soon.
One can be depressed under most definitions for CFS. Just most definitions exclude Major Depressive Disorders with melancholic features (possibly the severest form of depression). This is to try to help get consistent results in research so you are not comparing apples and oranges. The CDC empirical/Reeves definition allows people who just have depression to satisfy a CFS definition.Oh, and depression can be caused by CNS infection, so I would not rule out all depressive patients.
I don't think mixing people with CFS with people with depression (but not CFS) is the way to go for studying non-XMRV cases. I think studying the non-XMRV CFS cases on their own would be more productive.