 So, I'm going to present a very simplified perspective. It's been my observation at this meeting and a couple of others that there are at least three, three different main groups attending these meetings. There's the pharmacogenomics group, what I'll call the GWAS Common Disease Group, and then the more traditional medical genetics or rare disease group, and the last is the one that I come from. And talking about how you move things from research to clinical applications is part of routine medical care and standard of practice in genetics and genetic testing, and it has been the case for 50 years since the introduction of the karyotype and biochemical genetics testing both in the early 1960s. That evolved over time to become a more formalized process where we have CLIA-certified genetic testing labs. These are usually in a genetics environment or a pathology lab medicine environment. There's a critical team of experts that are involved in performing the clinical genetic testing, including lab directors that are either certified by the American Board of Medical Genetics and Genomics. I forgot its new name for that professional organization, or by College of American Pathologists and the Subgroup of American Association of Molecular Pathologists. There's usually a medical director or medical consultant associated with the labs that make cosine reports and is responsible and also in advising and developing new tests. And there may be other clinical experts, specialists associated with the lab. And for most genetic testing, genetic counselors are involved directly with the laboratory in an expert role advising about new genetic test development, as well as interpretation and reporting of test results. So this is clinical practice. There is a process of how we move new technology and new scientific information into clinical practice, and I've made a simple diagram here as a pointer. Okay, on the left shows new technology, much of it coming from industry sources, new information from scientific literature and meetings, et cetera. Many clinical labs have an R&D component, either formally or informally of variable sizes that's involved in implementation of new technology and also new information and development of new tests. Again, this team of experts associated with the clinical lab is the one that's advisory about what is, quote, ready for prime time and can be applied clinically, and CLIA specifies the clinical validation process for offering these new tests. Now, when the first one or two or three labs start implementing new technology or new specific genetic tests, they often report case series in the literature. So there's a feedback from the clinical labs who are participating in clinical research that contributes to the literature, either the laboratories and genetics experts or professional societies made to a meta-analysis, and then depending on the nature of the magnitude of the change, a new technology or a new clinical application, professional societies, usually the same major ones in terms of laboratory guidelines, that would be the American College of Medical Genetics and Genomics or the College of American Pathologist. For practice guidelines, the clinical specialty area would be the ones who would determine the appropriate clinical application of new technology or new tests. So from my perspective, what's happened now is we're moving into a major new technology change in capability with whole genome sequencing. For me, it's not totally distinct from what I started out doing with karyotype analysis, which is looking at the whole genome in a single laboratory test. And when we first started doing karyotyping, there were variants of uncertain clinical significance that required a large amount of empiric data and experience to determine which were clinically important and which were benign population variants. So now if we just move to genomic testing, I would make the argument that whole genome sequencing for undiagnosed patients is no different than what Genetics has adjusted to and evolved around new technology with every major technology increase. The magnitude, the frequency of variants of uncertain clinical significance or unclassifiable variants is much greater than in the past. And the opportunity to identify incidental findings that may or may not be clinically relevant and appropriate to reveal is much higher but not qualitatively different than what we've dealt with in genetics for many, many years. The question I would pose to the group, and I'm not sure if the answer is does this model, it has existing methods of assessing new technology at the individual lab level, forming networks or consortium and then having professional societies review the data for readiness for prime time. Does that work for pharmacogenomics and for risk assessment and common disease GWAS? Or do certain elements of it work? Can it be expanded? Or is the common disease risk assessment, pharmacogenomics applications of genetic data totally unique from what we've been doing in medical genetics? And I won't presume to know the answer for that. Now I would say that what we've not done a good job in genetics and has come up a lot at this meeting and other meetings recently is the determination of clinical value. So we've done a good job of the early translation feasibility clinical validity step. But in terms of the impact in health care, the impact on outcomes and value to health care system and quickly moving this into routine clinical practice in the broadest way we've not addressed. And so I think adding significantly more efforts in the genetics community and related areas in that area is important. And that's the last of my slide and the last of my comments. So I think that one thing that strikes me about this and in terms of the straddlers and I think is a huge issue that wasn't addressed is the training of geneticists and bringing up geneticists through the pipeline. And so the number of physicians going into genetics as we know dropped and has been relatively stable recently but did go down and has not really come back up. And I think that the physician workforce that's going to be needed for this is really lacking in a big way. I mean there weren't enough to begin with and there certainly aren't that many people going in, aren't enough people going into it. And I know this is something that the ACMG and we've certainly addressed at that level, but I really think that we need to think beyond this. How are we going to bring trainees? How are we going to bring physicians? How are we going to educate people who are really interested in translating this information, educating both practitioners and physicians? I think that's a big issue. Those are the straddlers and it's also true of generally a physician scientist that there are fewer and fewer people becoming physician scientists. I just want to point that out because I think that's an unimportant issue. I think that's an important point but it's a daunting task giving the declining number of physician geneticists being trained currently. I think the other opportunities look at other current geneticists and revamping their training. We had a discussion at dinner last night about genetic counselors being trained with more bioinformatics and genomics training and the opportunity to greatly increase the training program sizes for genetic counselors that might help in this area. Again, my bias as a PhD laboratory person, there are a gazillion underemployed PhDs in the U.S. that could be rebooted and trained in more clinically-oriented, computationally-oriented genomics-oriented way to serve a valuable role. So I'm going to point this out because I just got three emails about this this morning. Genetic counselors can't build. Well, but that's a problem that we should be addressing and not. But the genetic counselors associated with most laboratory programs are well-absorbed by the profit margins of the laboratory. Just a comment, David, to follow up. I like your concept of the R&D unit for laboratories. I would extend it to include some very well-defined clinical environments with the full complement of the workforce that is part of this as opposed to just the laboratory. I think what we've been talking about here and where I think some of the resistance is to implement things on a system-wide basis from the get-go. Why not think about having small centers of innovation in R&D within health systems that implement, innovate, and either diffuse or kill something when there's sufficient justification to do so? And so the concept, I think, far from the cancer community is the rapid learning healthcare environment. And I think it's not hard to do on a small and well-circumstribed basis if you try to do it system-wide. It probably is doomed to failure. And I forgot to comment. I've also borrowed from the cancer community the notion of a comprehensive genomics center or program that had all of the key expertise in one place to participate in the R&D validation and then implementation of new technology and new information. Going back to Kathy's point, there's also a movement from my understanding on ACMG and EBMG to move geneticists into the model of ophthalmology, where they do one year of internship and then go straight into genetics. And I think that's exactly the wrong way to go if you're talking about genomics. In genomics, you need to be broad. You need to understand heart disease. And you need to understand pediatrics. And you need to understand pharmacogenetics. And trying to get a group that's so specialized to genetics too early is only going to harm the field in the long term. Maybe Mike knows more about current trends than the restructuring training read. Sorry. So there are American Board of Internal Medicine or other similar groups in trying to think of something like a one year certificate of addict qualification or those sorts of things for let's say oncologist or adult cardiologist or these sorts of things to kind of potentially take up the slack with regard to qualified certified clinical practitioners. But I think that if you're sort of in the trenches as a medical geneticist, you see what's happening. And there are fewer people. And there are all these issues. And I think that this group really wants to do it. I have the MD clinical geneticists here. I just put them after the lab. No, no, no, no. It wasn't that. It's not that they're, it's just that we need more of them. Absolutely. I think as a community of how we get more people appropriately trained, that's all I'm bringing up. So David, there are already CLIA labs for next generation sequencing. Has the group here been working on standards in addition to CLIA? So where are you guys? There's a group of CLIA labs that have informally gotten together. I think the first meeting was at Harvard Partners. And Heidi Rimm organized that meeting. I'm not sure what their plans for continuing to meet. And I know that there's some working group discussion by the college about guidelines around next gen sequencing technology applied to diagnosis. At the moment, I don't know how much of that is gene panels versus exome sequencing. But yes, that group of experts doing sequencing diagnostics is meeting and discussing the same topics we're discussing.