I am not sure there is really a large enough corpus of data out there that covers multiple facets of this system to really give us a strong predictive value. Ju

I am not sure there is really a large enough corpus of data out there that covers multiple facets of this system to really give us a strong predictive value. Just variant calls is probably not enough. There might be, if we can somehow consolidate and integrate disparate datasets from various publications and labs. But I don't think we are at that stage yet. I am all for them trying though. I just don't think we are at a point where we can make a good diagnosis/conclusion yet.