 So, if I can just extract one thing from this in terms of a take away from NHGRI since Netta is sitting right behind me, it sounds to me that there would be some opportunities to potentially do a policy analysis to say, you know, are there things that we could do from a policy perspective that would reinforce the effort. So I just put that on a list of possible things to consider. I'll just also do a time check where we've got about 10 minutes. Right. So we still have about 10 minutes, you guys, for open questions and comments at this point in time. I think we kind of moved into some of the general discussion bulleted points that are listed on page 16, particularly some of the barriers, but I do want to open it up to other questions. I see Erin, you have one, and also Helen, I know you had one from earlier. No? It's answered. Okay. So Erin. I just wanted to come back to something that came up in Nazine's talk. When she was talking about categorizing variants in two or three and reporting them as no management, I'm just wondering how this gets updated and is there messaging that goes back to the physicians? But thinking about that on a larger scale with the ClinVar database and other EMR systems and clinical decision support tools, how does this ongoing curation happen? How does it get pushed out to patients? And when we do simplify for physicians and patients, how do we fix what might have been somewhat incorrect messaging originally? So there's a big difference between no management, you're off the hook, and while we're not really sure, so we're going to act conservatively for right now. Yes, so I should say that I'm in a sort of position that's helpful in that regard and also speaks to that integration between clinical and research, which I think is going to be really vitally important in having that two-way process. So as a clinician as having a head of a service to which I can, to a certain extent, determine how we're going to be organizing that practice and reviewing it. But also in my research, we're fortunate to have a national study to which we can recruit all breast cancer patients. So any data that I have in class 2 or class 3, I can, and we do, recruit all those people to our research because we want them to find new genes. But it also means that we have a mechanism there whereby we can still highlight all of those variants that we can come back to. So I appreciate that in that particular area in breast cancer research, and my play gives me a fallback so that if we reclassify, so all of the class 3 gets reviewed every three months. And if there's any new data that comes out in the literature, we can review that. But we also have a log then of all the people that we've previously sent that out to so that we can revise that. I mean, I do think that it is that balance between what, if we want to get it out into mainstream medicine, and it's a balance for ourselves as researchers or clinicians, between that balance of wanting to be able to use something which involves having to be pragmatic. We have to be pragmatic, and it's really quite painful. But unless we do that, we can't move forward. And also this issue that part of the reason I'm hardline with respect to the BRCA genes in terms of that information is that despite our best efforts, if data on variants which 99% or whatever is likely to not be pathogenic, if that gets into the patient record, people start acting on it. They start acting on it if they're guilty. They say, on the safe side, let's do this, or we'll do a bit of predictive testing. And so there, the potential for harm is higher than the potential for good. So that's that balance. And I should also say in terms of, because of breast cancer, each gene will have to have it, and each disease will have to be individually thought of. But if there is a very strong family history where we could really do those, of course, we just find truncating mutations, because that's what they look like. These mutations are in families with two, three, four cases. They're incredibly common. And there's abundant evidence showing that BRCA are not the cause of those. So that's the balance. I haven't got all the answers. A lot of what we're doing is a pragmatic way of dealing with the situation that we have in hand. Marjus, did you have a quick comment? Yeah, just in this country, since we don't have a national system or national identifier, it's obviously much more challenging. But Harvard Partners has developed a process by which there would be a service layer that would interact between the clinician ordering the test and the electronic health record where that patient is identified, and the laboratory that's actually running the test where if there is new information around a variant that's identified, that there actually is a maintained link that can then go back and message the clinician and say, we have additional information about this individual. We would suggest that we're reclassifying, and you can contact your patient with this information. I think the health heritage model that Bill talked about also has the potential to do that, but to do it on that patient-managed side of things. So we're going to have to come up with some solutions where we can confidently identify patients who have had tests where we, when we do have new knowledge, we can get that knowledge back to them somehow. It gets, it rubs up against the, you know, the recontact and that. But I think in the long run, given that this is an evolving field, we have to solve the problem. So I'll just add to that two quick points to think about the payer standpoint. Since there are some payers in the room we've been talking about that, how does that additional analysis and messaging get covered from a payer standpoint? And in the era when we have a whole genome sequence sitting there that we're going to reaccess for new tests or new analyses in the future related to different diseases, how does that additional analysis get regulated by CLIA since the analysis will be very discreet from the sequence generation? So I can address that. The system Mark was mentioning is the gene insight system that we developed. And one of the reasons we developed a mechanism for this automated updating process is because these services are not billable today. We have no way as a clinical lab to bill for a reinterpretation and that's very expensive. And so the system essentially works in an automated way so that if I sign out a new case with a variant that was reported in an old case, any of those old cases automatically get updated at the same time without me doing extra work. And that was a way to get around the costly process of doing that. The other thing that we've done with that software is taken it to the FDA and it's a Class 1 exempt medical device in terms of how it functions because it does act in an automated way and therefore it's under FDA's purview to sort of regulate it. But that also enabled it to be a lot more robust in terms of how it functions within the healthcare environment. And we are now integrating that directly into our EHR system at Partners Healthcare and working with Mark about doing that in our mountain health as well. So it enables some process but it doesn't cover everything because we don't always see a variant again and some physicians may want us to go back on their old case and redo that and so I still think there is the need to develop a mechanism to bill for simply evaluating data again. That's definitely needed. Howard. So my thoughts are along the same lines and I'll start by reiterating something I said yesterday morning but clarifying it that while clinicians want a clear, fast, concise answer they again don't want to be lied to, don't want to be told something's negative when in fact it's uncertain and that does breed ill will and clinicians typically tune out a channel that tends to give useless or wrong information and I think there's a preponderance of evidence that shows that clinicians actually are capable of dealing with uncertainty as long as it's clearly and concisely stated as uncertain just don't know what to do. I also have a little concern therefore about reporting your Category 3 of possibly assigned as negative. In my mind I equate that to Jonathan's bin 2A or maybe 2B. What I would propose is report those out as they are uncertain and I'll further go on and make perhaps an even more disruptive comment or suggestion and that is that the bin 1s, the clearly abnormal, this is disease causing are probably among the easier things to manage and when we think about the resources of medical geneticists this is going to break down silo walls but maybe medical geneticists are no longer so critical for many of those. We've already seen that in CF and sickle cell disease. How many clinical geneticists in this room have seen a CF or sickle cell patient for ongoing care? I see it happening in neurofibromatosis now as well. So perhaps the role of the limited medical geneticist specialty resources shouldn't be in the clearly normal and clearly abnormal realm but in these possibly assigned these variants of likely but not definite significance and that's where we can bring to bear our skills and talents as medical geneticists. Can I just clarify in terms of that negative? The management report is not saying that the report of the sequence variance is negative. There are two different reports. There's a report of the outcome of the genetic test and that could have all of your synonymous, it could have whatever you like. That is just the report of the outside of the test. The management report is how are you going to manage that patient and I would very strongly argue that if you have a mis-sense variant that you should not be managing that as a positive BRCA mutation. You should not be suggesting to that patient that they should have prophylactic mastectomies or uforectomies or doing predictive. So in that sense it's how you're managing them. So the report that comes out as negative is you are not managing that as a positive BRCA mutation and I very strongly believe that that is the case and there's abundant evidence that you shouldn't do that. So that's why I think there's a difference between what's negative and positive. In terms of your sequence variant report, that's an information, that's the test. It has a whole load of information about the sequence information. In terms of your clinical management, that is potentially different and I think that is. And I totally agree in terms of those, there are two things that I agree. In terms of, a number of people have said, I've got my patient today. She's in my waiting room today. What am I doing today? And that's what that clinical management report has there. Then there's the whole aspect of what is all this interpretation and the change of potentially changing that for tomorrow and how you're going to interpret those sequence variants. And I don't think generally, if you've got a busy clinic and you're seeing things that people want and are able to deal with those nuances there, they need it to have been stratified down for the patient that I'm seeing in five minutes today. So I think that does address that issue. So there's just one nuance that I didn't articulate well enough. And that is that while you're absolutely correct that it's wrong to manage the BIN3 or BIN2A as positive, I think it can be argued that it's also wrong to manage it as totally negative. And I would disagree with your thesis that we must be discreet as yes or no, manage it as high risk or not high risk. I would propose that there's an intermediary zone that is very great, very muddy, very hard to define, but not clearly normal, not clearly positive. Well, that might be for some cases and the number of stratifications you have depends on your disease. So if you've got, there may be three, four different ways because you can do this screening for that and that, you have to tailor it. We sort of spent a long time about this and he's thinking, how do you manage an uncertain? What is one? What are you going to do? Take off one of their breasts? I mean, it's, you know... Well, yes. I'm going to stop this. And the bottom line is that we do this all the time with guidelines. Guidelines can, you know, we follow or we don't follow a guideline. That's what it is. And we can choose using clinical adjustment to follow or not follow. We don't see incompatibility with what we're doing with these two things. All right. But I do think we're, if you want me to write time. So thank you. Thanks to panelists for coming and speak today. I'm going to... Mark, do we have directions for lunch or...?