 Yeah, so I was asked to give an update, a status report on our clinical sequencing exploratory research program, which we've come to call CESAR, but I may wander around a little bit on my way there. Just some historical slides. You know, it's only 10 years ago that we finished the human genome reference. And following that, Francis and Eric and Mark Geyer and others published a vision for the future genomics research, which included this, quote, the technological leaps that seem so far as to be almost fictional, but which, if they could be achieved, would revolutionize biomedical research and clinical practice, for example, ability to sequence DNA at cost four or five orders of magnitude lower. Of course, this was the drive to the $1,000 genome. So this was just 10 years ago. And of course, we, at NHGRI, on our genome.gov site, maintain this graph you've all seen, probably, the cost per sequence genome versus Moore's Law. And this inflection point occurred in mid-2007, where the next-gen sequencers came online. And we suddenly took a slope that was following Moore's Law and it plummeted. This graph that I have on here ended July, two years ago. And now we've reached the point in costs where an exome, a whole exome, is $1,000. And whole genomes now can be purchased with a prescription from your doctor for $4,500 from Illumina. And it hasn't ended. We've kind of flattened out a bit with the current technology. But there are newer technologies coming. Oxford Nano made quite a splash last year with their thumb drive sequencer. This is a working prototype that you put a drop of blood or whatever into this hole. And through nanopore sequencing, the sequence goes right into your laptop, powered by your USB port. Look for more news from this group, I think, in the next year or so. So genomic sequencing is here for clinical use and we've got to figure out how to do it. So we've been looking at, of course, at NHGRI, this path to genomic medicine. And what do we do with routine genome sequencing coming online? And this, facing all the issues that we've been discussing, led to us to start the initiative for what we call clinical sequencing exploratory research. But briefly, before I go there, of course, with all the sequences, also all the data slide I borrowed from Eric, the drinking from the fire hose phenomenon is coming and Terry or Lucia commented that we should mention, it's nowhere else on the agenda, the new initiative at NHGRI on big data. So we're all aware of the big data issue that's going to get worse. And so at NIH, a while back, and these are slides from Eric, big data to knowledge initiative was started. NIH is trying to get on top of this difficult problem. By the end of the decade at NIH, we like to enable a quantum leap in the ability of the research community to maximize the value of the growing volume and complexity of biomedical data. And this, everybody's jumped on board at NIH, a working group was developed last year with 125 members and almost all the NIH institutes and centers are part of this new initiative. And just quickly, it contains four programmatic areas, a major one in facilitating the broad use of biomedical big data. Some of the things we were talking about yesterday, how do we share all the genomic and phenotypic data that will be generated in the clinic and more broadly, developing and disseminating analysis methods and software for big data, training and an effort in innovation in what we like to call centers of excellence to innovate in this area. Where it is, we're in the program is in the workshop phase, a number of workshops, I mentioned one in particular coming up in September, enabling research use of clinical data relevant if you receive an invitation, I hope you'll consider participating in this important workshop. And so with the workshops going on this year, then funding will start in fiscal year 2014 with a model of actually expanded funding leading through the end of the decade. So pretty serious effort and gotten pretty well organized at NIH. So I just wanted to mention that and now I'll turn back to clinical sequencing. Obviously this was set to alter the clinical landscape. We're not talking about genotyping. This is broad genomic sequencing with all the ramifications for incidental findings and other things that go with that. So we developed this program, CSER, to research the challenges to applying comprehensive genomic sequencing of the care of patients, how to generate the generation and application of the genomic sequence data in the clinical workflow and timeline, how to interpret the data for the physician and communicate these results, the results to the patient. And what was a little bit innovative in this program was we really wanted the ethics, the psychosocial implications to be a big part of each of these projects to be built in and be fully integrated with each of these projects. This is, if you've seen the heat map, this program is really embedded in the fourth domain for the NHGRI strategic plan advancing the science of medicine really at this cutting edge of where we are today in trying to move genomics into medicine. I point this out partly because I also do a lot of work in understanding the biology of disease and this program, CSER, is really about understanding and researching process and the psychosocial implications rather than discovery of new variants and associations. So each CSER project has three sub-projects. Each of the sub-projects has senior leadership. Project one is the clinical study, the choice for what clinical question you'll be addressing. Project two is the actual sequencing analysis and informatics that goes on all the way from variant calling to decision support tools. And project three is the important ethical and psychosocial implications. As I mentioned, we have each project required strong leadership in all these areas and then important management structure to coordinate this. So these are the current grantees in CSER. I point out the PIs and also the LC lead for each University of North Carolina. Jim Evans' name came up several times yesterday with Gail Henderson doing exome sequencing for several different disorders in North Carolina. Levi Garroway and Steve Jaffe lead a project at Dana-Farber looking at solid tumor sequencing. The expanse of psychosocial and ethical questions in the projects are quite vast. The concerns of somebody coming in with colon cancer, with acute disease and returning results or incidental findings is much different than a relatively healthy patient. Robert Green at Brigham and Women's with Amy McGuire and Baylor have our only study that's doing whole genome sequencing. All the others are exome. They have a part of their cohort is actually healthy patients. They're recruiting to do whole genome sequencing in some longitudinal studies. University of Washington with Gail Jarvik, Wiley Burke, Malia Fullerton is looking at colon cancer, colon polyposis. This is not tumor sequencing but actually looking at familial risk for colon cancer. CHOP with Ian Crantz and Nancy Spinner looking at pediatrics, a number of disorders, hearing loss, developmental delay, a couple of others, a variety of disorders and also tackling those important ethical questions around pediatric genomic analysis. And Baylor College of Medicine led by Sharon Plon and Will Parsons doing pediatric tumor sequencing. So quite a diverse group. We reissued the RFA actually last year and are about to add a few more groups to this to kind of expand the consortium. We work very closely at NHGRI in this consortium with a separate group, the return results consortium out of our ELSI group. These are our grants variety, a lot of the big names in ELSI research and we actually meet at least once a year together and obviously tackle a lot of the same issues. A lot of the work coming out of CSER is actually going to come from our working groups. These are very important to us. We want to develop recommendations, survey different approaches for clinical sequencing. So the dissemination of this will come through these working groups just real quickly, phenotypes, how to standardize some phenotypes and be able to get some interoperability in the phenotype area. Sequencing standards by Levi Garowitz. We mentioned this this morning about in standards. The interesting thing they're doing is we have five groups doing exome sequencing and they're surveying coverage. There are, for example, the 57 genes recommended for return results, for instance, findings by ACMG. Several of those genes are actually very poorly covered in the exome capture, at least that region, usually the five prime region is very poorly covered in the exome capture methods. So we're trying to understand this and develop some better approaches to get at least thorough coverage of those genes that we think are important. Constant questions around actionability. This group actually has a paper submitted that just kind of surveys the different approaches in CSER that are being taken to make judgments about clinical actionability. EMRs, how to get these data into the EMRs, psychosocial measures. Still working on informed consent. Never solved a pediatrics and this is another emperor. Clever take on the CSER name, but this is a return, another return of results group. So this is where a lot of the products from CSER will be coming from. Just to give you a little update. These are now a little over a year into their grants and do the all six awards. There have been 455 people consented as patients, participants. 170 have been sequenced as of last month and results are beginning to go back to these patients, participants. Of course, a lot of this also is the physicians are very important participants in these studies. We're studying physician preferences, patient preferences, really studying in an empirical manner the interactions that occur and the preferences that occur as we return sequencing results. A lot of physicians also have already been recruited. Just the last few slides just to give you a little more flavor for some of the questions CSER is tackling on the incidental findings area. All reporting incidental findings, half included. Incidental findings automatically go into their primary indication report and then half of the groups actually have a diagnostic report and then a separate incidental finding report. Most sites allow opt out. As I mentioned yesterday, most patients do not opt out of the medically actionable incidental findings. Some cases it's not even an option in these studies. All allow non-medically actual incidental findings for the patients to opt out except for Brigham and Women's which part of their study is actually to do healthy patients and hold genome sequencing. So what categories of incidental findings? The full gamut, disease risk, carrier status, pharmacogenomic variants. You can't actually get blood group from exome sequencing but you can from whole genome. What types of disease risks are returned? Some of our groups have a predefined gene list for what they will return to patients in terms of disease risk. And then others are using or actually doing this on a case-by-case basis. And a kind of, I would say, non-scalable board type approach to make decision patient by patient. All these groups do have one of these multi-disciplinary committees to make decisions about actionability and what to return again. Some do it case-by-case. Baylor has really kind of taken a lead here. The Baylor College of Medicine led by Sharon Plon. And they're now doing over 200 people in their clinical sequencing group each month. And then others have this all-priority categorization. We haven't really touched fully on this. The variant classifications intend to return pathogenic and variants of unknown significance for the primary indication for the patient. And then pathogenic variants for incidental findings. But then a big challenge is how do you determine this? We don't really understand the biology fully yet, certainly. So what's sufficient evidence for pathogenicity? And of course, all these groups, everybody dives into the literature and the databases as they are. And they find things like reported as pathogenic, segregates with disease in a family. Or in some cases, a new variant will have no literature record, of course. And then how do you predict potential pathogenicity? So this is a huge challenge that's being tackled now. CRVR, we hope will help. But a lot of this really is going to take, as we were just talking about, the evidence generation and even some research on the functional aspects of new variants. Okay, with that, I'll stop. And if there are any questions about the CSER program or any of this, I'm happy to take them. Yeah, Pearl? Have you ever talked or asked the CSER participants what they would see as a national policy? And I just think they've been playing in this field. It might be an interesting, you know, to bounce back on some of these groups, kind of what we're struggling with. No, I mean, they tackle things like CLIA, all the groups that developed CLIA and dealt with that. And now, talking more about reimbursement for some of the assays and that sort of thing, but no grand strategy. Yeah, that's a good idea. Yeah, Lucie? So one thing I might just add is as part of CSER, we just within the past couple months funded a coordinating center. So that scaled Jarvik at the University of Washington. And we did that in part because we recognized that although clinical sequencing is being done individually at institutions, I think we have an opportunity with CSER to sort of get at some of these larger issues and figure out what lessons these groups have learned and best practices that we can share. And I think the coordinating center is going to have a key role in doing that as well as any sort of education and outreach to professional societies. I mean, we just had our steering committee meeting last week. And I think we had a very lively discussion of the ACMG recommendations, among other things. And that consortium of people where many hats including representation in ACMG and pending efforts and clinically relevant variants and so forth. So I think there's going to be much yet to come. Okay, Karen? Yeah, you mentioned the meeting in September for research. Can you go back to that slide and just... Maybe. Yeah, I'm not a big data expert, but maybe somebody here is. Enabling research use of clinical data in September. So that's just getting organized and I think chairs have been selected and they're just formulating their plan. I don't know. Anybody know? Did that help? Yeah. Okay, Karen? Yeah. This is a difficult question. So I'm going to ask it to Brad and I also want to ask it to Les. And that's the relationship between the research that we're doing and the process that ACMG went through. Because it looks like you're asking a lot of questions that may have some relevance to ACMG's decision making. And you had some of those people on the committee. And I have a worry that there may be some unintended consequences of pushing ahead. I understand the need for the list or the charge of a list, but there are some big leaps that were made here. And we have some research that is being done that may be asking some of those questions. So it's a difficult question to respond to, but I didn't know if either Brad or Les or whoever else was here. And as you pointed out, some of the people that were in the deliberation are also doing the research. I can speak to that point, I think. We certainly recognize that there are efforts underway to gain data that will help to answer these questions. And the truth is that this is being used today as we speak clinically. And people aren't waiting for these studies to begin to apply this. And so there was a strong feeling that there needed to be a flag planted somewhere to give a standard and guidelines with full recognition that we are going to learn more as time passes and likely need to revisit various aspects of these guidelines right of that new data. But you can't wait what could end up being years, perhaps, for the outcomes of these studies when clinicians are using them right now. I just thought, did you consider whether it would be reasonable or even unethical for clinicians together with patients to choose to opt out? So they would just want certain information, but not all that you have on your lab test, because that's happening. And that's actually coming out of the discussions with some of the Caesars. And it's an interesting research question. But I just wondered if you actually, obviously, made a decision about that. You set a standard, but it's happening. So I was wondering your views on that or Brad, if you want to chime in. So you're asking if we considered the possibility that people would choose not to? That the clinician, it gets to Pearl's point earlier, that this role of the clinician together with the patient as a couple. And these would be the role of the lab and the ordering. And the patient that just says, you know, I came here for a reason. I'm not interested in that. And the answer is, well, then what? Then you don't get this. Mark, yeah. Yeah, so I think there are several points. And obviously that was a, we discussed this quite a lot. And I think the things that I identify as being, that bear on that the most appropriate lawyer, I think people are confusing the idea that somehow whole genome sequencing is a standard of care. And that if we don't offer whole genome sequencing with people, you know, having the option to pick and choose, we're somehow withholding something vital to their health and well-being. And the reality is, as we've been talking about over the last day and a half, is that this is, you know, emergent. And as with many things, it's emerging into practice probably well ahead of, you know, what is going to ultimately be best practice. So that was one contention. The second one is that there is a body of literature that shows that asking people to make choices about opting out in a hypothetical situation based on presenting hypotheticals of if we were to find this, that, or the other thing. That when people are faced with the real situation as opposed to the hypothetical situation, they choose very differently than they've done. And so we thought that it would be presumptuous to assume that people could really, truly make an informed condition in the pre-test setting to do an opt out. Again, the point that is important and relevant here is that, as Les pointed out, the result comes back to the clinician. And we expect that there will be a discussion between the clinician and the patient at that point as to whether or not results that have been returned from the laboratory to the clinician will in fact be returned to the patient. And that we think is where that conversation is most appropriate to take place because that really reflects what is done in other medical practices. So to reinforce what Les has said, what we're really trying to do is to move genetics from our exceptional position into something that is much more concordant with standard medical practice. And so this was, you know, so the bottom line is as we spent a lot of time on this, those, for me, were the arguments that were the most compelling as to why we elected not to do a pre-test opt out. And my position in our clinical research approach is that if people are uncomfortable with the idea of incidental findings, that the autonomy is don't do this test. We will do everything else in our power to use alternative methods to make a diagnosis. But if you do not want this information, then this is not the test for you. One of the things that was interesting to me last week in the discussions was this isn't going to be an overnight shift in the clinical labs. You know, there's actually going to be some slow progress made as they think about how to implement the ACMG recommendations and how best to deploy their resources in doing this and make sure they do any responsibility in a responsible way. So both the practical aspects of interrogating these 57 genes, but also working with their IRBs and how to do it. I'd like to sort of just advocate one more time for the idea of sort of creating very large data sets of variance coupled to clinical data, because I think that's really going to be the only way to deal with. The one-offs I don't think we're ever going to get to unless you understand the biology, but the people who have the 0.5% variant that we just don't know what to do with, I think the only way forward is the phenotype dataset. And like Les said, the phenotype is what really rules there. So finding out what happens to people with these BRCA1 or MYH7 variants over time. Is it, you know, the frustrating thing is that they exist. Myriad could do it. GeneDX can do it if they put some effort into it, or somebody can partner with them or something. Okay, I think we need to move on. Thanks, Brad. Last presentation before break is going to be Linda Essler.