I think for the first point – and thanks, Doug, for asking this important one – so we understand the biology of cell-free DNA very, very well. The kind of biomarkers that are associated with cell-free DNA. We have studied the science and biology of it very, very well. So I think we can kind of distinguish some science fiction, which doesn't, like, biology – there is no biology for it versus the reality of what different technologies can do as long as they are looking at cell-free DNA kind of biomarkers that, again, we know the biology very well. In science, we can never say it's mission impossible, but I can tell you, like, I think some of the scenarios you mentioned cannot happen really in a probabilistic way, like with very, very high probability. So, like, stage two, three, four is going to be better with cell-free DNA than stage one. And the performance metrics, as long as the team is capable, frankly, that kind of a response curve is what cell-free DNA can do with some kind of upgrades that I think hopefully we are going to see with our Shield V2. But response curve would be similar. AA, I think, is a different topic. If somebody thinks they can have just a 70% plus or 70% and change CRC, but they can detect materially more AA, we just let them learn some lessons through biology. In terms of, I think, profitability target, I'm more confident than ever about the targets that we shared in our Investor Day last fall. So, 1 million sample, $500 million revenue with a $200 COGS at a time. But there are some stuff that went better than our base case, like this first line indication is better. The way we are doing the ADLT pricing is higher than before. But it's still, again, too early for us to adjust any of the expectations frankly. We cannot get ahead of our skis. We will not get ahead of our skis. We just need to see how the volume ramp is going to look like, how the payer mix is going to look like as we go through this launch and the first few months after. And then, maybe we can talk about long-term perspectives with better data.