 Great, I think we can open it up for group discussion and then maybe just to kick things off I actually had a question for Robert actually based on his and we had a little sidebar on this in terms of control groups in a randomized Study and so forth He presented some nice examples of how Caesar has been carrying this out in general About four different sites and so forth But I wanted to hear your thoughts about how to think about this in terms of cancer sequencing in terms of including control arms and In some cases, I think considering some of the ethical issues of potentially not returning a potentially actionable result In those patient groups Well, I I think I'll dodge that question because I'm not a cancer expert And I think part of the excitement of a cease or two would be to try to create some creative Solutions that sort of thing, but I will say that you know, I think that in cancer and in other conditions You have a lot of flexibility about the way you design clinical trials You can design them for time lag So for example, you can design an intervention and in a control group instead of not having that intervention You could have it later on in time You can design different dosages of a particular intervention So for example instead of a I don't know a drug that's tailored to a particular cancer You can give one that's a little bit tailored and one that's even more tailored So there's a lot of creativity in the way that you can I think accrue health evidence that would be part of the challenge To investigators in my mind and in a cease or two Levi So I wanted to ask about a topic that Robert touched on in his talk and also Gail touched on in her talk But we haven't actually expressly discussed it and this is the issue of so Robert You're pointing out that one of the ways to approach Refenotyping in the setting of for example, very certain significance is sometimes one could go back and do either be for phenotyping or even sometimes, you know biochemical base phenotyping, etc And and Gail in her talk brought up this idea that there are groups around that are now doing assays, you know functional assays for variants And so this is something that raises a question that one could imagine emphasizing the the E in Caesar in 2.0 Which is that there there are contexts where there are variants of uncertain significance, but they're plausible and The context is right, but you don't actually know if they're functional But there there might be instances where one could with the right collaboration the right teams do kind of a Very focused functional assessment either of the variant itself or if there are Cells or tissues from the individual where one could do an additional kind of deeper functional screen So one it's sort of an appealing and novel and innovative Direction of one could go I guess maybe this is even a question for NHGRI Do you guys have the appetite for For that kind of approach there's certainly a lot of interest in there There's a big space out there when one where one could Explore the idea of more phenotyping more functionally based phenotyping for variants that are crying out to look like actionable We just don't have the space and we may never if we just wait for kind of the clinical experience to grow There may be a long lag before we Learn about that Go ahead So I think it's really that's a really important idea We've taken the approach at UVM that if we do research on the clinical variants that we're seeing that we don't know What they do we're much less more like much more likely to feed back clinically useful information with the Functional genomics research that's being done there And so I don't know if there are ways to link the variants that we don't understand through the Caesar project to functional genomics parts of NIH more broadly even Which also could get us to guidelines that actual clinicians believe in Because they'll be the variants that they're seeing in their patients and how to use them So I think I mean some of the Caesar sites are doing this in a very limited way and obviously there's broader interest My understanding is the UDN program also has a significant functional component as well But it's something that certainly would be of relevance to Caesar going forward especially in the context of The the workflows you already have something to explore. Yeah, and to more broadly answer Levi's question You know, it is something I think we are definitely discussing and it's gonna be part of the function You know, it's gonna be part of actually we already discuss as a topic for the next GM meeting is how to make connections with Functional and functional work and I think that you know I talked before about the virtuous cycle that came out of the discussion with the sequencing But I think there's another way you can think about the basic research of that virtual cycle To also think about things very much at the functional interface Dan Huge fan of this because in the iron channel world, it's you know The assays are easier than most or that many but but there's also this sort of limitation in the sense that you do have People who carry variants that are clearly functional at the in vitro level and have no phenotype and people and there are and there Also variants that we think clearly confer a phenotype But the assays that we have unless you sort of start to really drill down or make mice or whatever You you really miss the phenotype. So but I I'm 99.9 percent in favor of the idea that 100% favor of the idea that that that adding no no adding adding that phenotypic information is a is ought to be a helpful way of going forward in terms of interpreting the data just with the caveat that it's not Perfect, but it's but it's it would be a huge step forward. I think the iron channel should be done first. That's Goes without saying So I mean I completely agree that we need more functional analysis I think the question is how to more most effectively do it because I think one-off encounters with a patient leading to a one-off functional study May not be the most efficient nor effective way to do it and one of the biggest challenges in variant interpretation is interpreting literature Unfunctional analysis, especially when it's been done on one or two variants Limited validation of the actual assay to be a good predictor of of the actual phenotype, etc So I think one thing to do is to continue to share these variants or groups that have functional approaches that if given access to large sets of variants have actually been observed in patients and the ability to go back to those patients and Correlate their functional studies with phenotypes back and forth I think would be an incredibly rich exchange and so I think we just have to figure out how to Connect groups that are doing more high throughput Studies that are well validated assays with the patients and the data that they're coming from Steven Steve Jaffee from you pan different topic I want to give a shout out to some of the conceptual work that's been done in Caesar and emphasize the importance of that work in Any future work that comes out of this and specifically what I have in mind on one of Robert's first slides You talked about the different senses of clinical utility. You were Channeling your inquiry there You know sort of narrow sense broad sense and broader sense and one of the key questions We have is so which of those senses should we take into account when we're deciding what to pay for with collective funds? Whether they'd be taxpayer funds or health care premium funds, you know the broader sense anything that matters to somebody You know, is that the sort of thing that would sort of imply an obligation to pay for that with collective funds? We're in that spectrum should we be thinking about that and that's not entirely an empirical question And so that's just an example of the sort of conceptual work that I think we've been able to do in parallel with the empirical work That has gone on in Caesar, and I hope we'll be able to do in the future Robert did you want to address that? Yeah, no, I not that particular, but I just want to say building off of that that two of the real key Innovations I think in Caesar has been in the larger sense as Steve was pointing out integrating LC Instructively with the implementation work, and you know, this is not else. This is this is not your father's LC. This is not this is not Sort of staking out positions in a way that could be construed to Just sort of be slowing down the progress. This has been a real working interactive relationship that iteratively works back and forth between the data and the policy and the second thing is Dan Building off of your comment I think we have the opportunity to really Transform the entire notion of phenotyping from an 18th century phenotyping which is you know tapping on the knees and listening to the heart to a 21st or 22nd century phenotyping which is a true systems biology approach to the entire biological spectrum of disease and Caesar won't do that by itself or without a ton of help, but we could be one of the points of that particular spear so Again in my laboratory world The reason I'm doing these tests is is to help with the diagnosis Sometimes it's to confirm a diagnosis. Sometimes it's to roll out But in the clinical utility world Where does the value of the diagnosis come in to place even if there is No treatment there is not I mean, I don't know how to look at at outcome studies if there's not really a treatment but To come up and say here is the diagnosis therefore. We know there's no treatment You know at what point does that become valuable and one of the challenges with the genetic diseases is that the outcomes are Not always on the patient, but the outcomes are on the family members So how do you pull that together to demonstrate value? So I think and we'll hear from from that might after after lunch I think and he can give you a very personal example But from the perspective of genetic disease as a practitioner and also from the undiagnosed diseases network I think we place great value and simply solving the puzzle because not because solving the puzzle is the end But the solving the puzzle is the beginning and it's the beginning of the next search as to Learning about the biology of the disease and learning about the biology disease takes us directly towards therapy So I think there's huge value in that and thinking about sees of 2.0 Then I think that shooting definitely be part of it And one last comment also I think maybe someone was whispering over here But we do have much more power now with saturation with the dentist's type experiments and CRISPR to do much more High-through-put functional studies. They're not to be on end all but we definitely can that can take as much further in In terms of a disease and a dish along with stem cell technologies and to Robert's point We can measure people at times relatively cheaply in a way that we could not do just a few years ago So I think that would also be something to consider strongly. I just want to follow up on the comment someone made earlier I forgot who it was. I apologize But I'm wondering whether it would be useful and interesting to have more emphasis on Comparative effectiveness between what we think of as genomic testing and what we think of as the rest of medicine It's very surprising for example, how what the positive predictive value is of a cholesterol level That's really nowhere near what people think And I think there are many other things that we routinely do in medicine that are as Good or worse than what some of the positive predictive value is from genomic So maybe some comparative effectiveness between genomic and non-genomic medicine would be interesting Sure. Yeah, I just wanted to get back to the phenotype issue in the refenotyping. I think we just have to I Think in Cesar 2 it will be extremely important to have some Baseline phenotypes agreed upon in advance of something we didn't do in Cesar 1 because our patient populations are so different I think if we're going to refenotype people we have to be very careful at the variants we pick So just from the example that That Robert Gabe some of those are incredibly well Documented variants were frankly just phenotyping a few more Cesar patients won't be helpful Whereas there are others that are associated with very severe Severe inherited diseases where phenotyping those individuals would be very helpful And so I think there is a process of prioritizing the refenotyping that would really maximize Our knowledge of Cesar patients and not be redundant with other studies Katrina So I guess I was interested in the panel's comments One of the things that has really appealed to me about Cesar over its history has been I think often as we're seeking to improve Medicine let's say that we either focus on developing the perfect tool So we have the perfect test or we separate that and focus on how are we going to use it in clinical practice? But the reality is I think as you're suggesting so much of how good the test needs to be depends upon the clinical Decision that we're going to make with it and so much of this is really an iterative potential process And I guess I was interested as you all think but you maybe even your echo comment You know, how do we think about maybe the future of Cesar is continuing? I think in one of the few programs to really sit in that interface Recognizing that you are continually trying to make the test or the tool better But that's actually understanding the application of that in clinical practice that is Critical for us to be able to make that tool better because the perfect tool to be honest is never going to exist And we're always going to make trade-offs of where do we need to focus on making the tool better? And where are we actually at the point that it can already inform the decision we need to make in clinical practice? I just be interested I think Robert both you and you and commented on that But if you look forward do you see that as an opportunity for Cesar? What I mean, I think you said it very well I mean, I think we have to balance the two and I think actually we've been doing Well, we you because we're talking really about Caesar here So people in the room not part of Caesar have actually been doing a very good job of that I think paying some attention to the technical aspects improving that making sure that process is iterative And then finding new places to apply this to then testing those and feeding it back Actually, I see I see that as a major success of you know, as I was reading through the packet and obviously many of the Papers I've read and we're aware of the good works of the consortium But I think a Bob's point is also a really good one and it sort of reflects the point We made earlier that I think we can hold ourselves to too high a standard and that comparing to Standard tests that are currently used and in practice and accepted by other people will make the narrative easier But I just think is more realistic you you brought up the broader societal good as a measure of clinical utility versus Proving clinical utility at the individual patient level What what does this group think does this group think that it's worthwhile? Investing resources on the broader societal clinical utility given the state of our so called healthcare system I would just have to say I think we have to be skeptical of the notion of social proofing And there have been an awful lot of ideas that have been socially proved Or proven if you will that have turned out to simply be wrong and That's where the primacy of evidence Has to be maintained and I I think it's it's interesting to observe as things get taken up by Society in general But that doesn't always Result from a sensible uptake and five ten years later than the things that were embraced or found to be lacking Putting it a different way and I know I'm gonna get ribbed for this later. Ceaser is the conscience of the biotech Tsunami that is coming