 Yes? Okay, so it works. So I'll give a little preamble to how the clinical center and intramural collaborations could work, while Les Lodges' slide. Anyone who comes to the clinical center has to be on a protocol, a research protocol, and they don't pay anything. Sometimes they'll pay travel, but that's all they'll pay for. But the finances are a little bit important. Every one of the institutes that engages in the clinical center pays a tax. We call it a total tax, and then it doesn't have to pay for any of the infrastructure. So things like the labs and the radiology and rehab medicine, stuff, those things are all free. But the collaborator, the principal investigator, does have to pay for actually taking care of the patient. In other words, have a doctor or a nurse practitioner, stuff like that. So that's an important thing to know when you want to establish a collaboration. So the types of collaborations that can occur are, you call up an intramural person and say, I want to work on this disease. We work with me, and maybe you can see the patients at the clinical center. That's one. Second would be a bench-to-bedside grant, which are now given priority to collaborations that include an extramural investigator. Those pay $135,000 a year for two years, and they can, that the money is divided up however you write the grant for. Now this year there are like 120 applicants for roughly about 10, or maybe as many as 15, depending upon how the ICs chip in extra money. In addition, there's going to be, there's in the process, a formalization of opening up the clinical center to extramural investigators. And that will involve grants. And I don't know the exact amount, but it's more than the $270,000 for the bench-to-bedside. But those really, they have to work out the details of mixing intramural and extramural monies. It's a color of money issue, but they are working on that. And then finally, NCATs could change things considerably. I don't know exactly how, but I think it's definitely in the horizon. How's that? Hopefully positive. Indeed. I guess depending on who the director is. So, and then I guess the other, the final thing I'd mention is how the Undiagnosed Disease Program can be engaged in collaborations. And I think there are two ways. First of all, we have found genes whose function we don't know. So if they're experts at your institutions who know a lot about the function of some of these genes, we'd be happy to sort of tell you all about the clinical stuff. Because these patients are incredibly well phenotyped, and we're pretty sure about the genes if we're going to, you know, get them away. And the second thing is that in the whole exome sequencing analysis that we've done, it has required us to do some creation of software for filtering the variants. And so we have developed filters for analyzing small families, Mendelian inheritance, and now we're working on one of the issues which we think is paramount, and that is misalignment and how to address the misalignment issues, which are essentially recapitulated every time you do a whole exome sequence. Okay, that's all I've got to say. Questions? Yeah, if there are questions, I'm happy to answer anything. So quickly I just tell you one little story that will illustrate, I think, an approach that we haven't talked about a lot here today, which emanates from rare disease genetics, but then goes into more a common issue that I think a lot of us have. We, like many people are doing what I'm now calling the standard clinical rare disease genomics drill, which is using exomes to solve rare Mendelians. One that we did was a disease we thought was a rare pediatric metabolic disease called combined malonic and methyl malonic acidemia with severe acidosis, coma often strokes and infarcts. Did the exome trio drill, found a variant potential causative gene, analyzed a bunch of additional patients, found many rare mutations in them, localization, functional correction, et cetera, and sort of solved that disease. Then we were interested in using our clincy cohort, which is a cohort of middle-aged, I like to call middle-age, young middle-age people, to ascertain carrier frequency. We thought it would be good to know the frequency of this trait. And actually, the minor allele frequency was significantly higher than we expected to find. Our presumption was, well, we must have overcalled the variants in this cohort because the disease frequency that was about five to ten times what we expected it to be. What the data looked like when we pull up our clincy cohort, which now concludes nearly 600 exomes in patients. Again, phenotype for cardiovascular disease was the following, and what really caught our attention was this result, which was at one of our nearly 600 patients. This iteration is actually about five or 600. Patients was homozygous for a variant, which, of course, the obvious conclusion is, again, we have overcalled the variant, then it's benign and not causative. However, we have the ability in this cohort to do iterative phenotyping, which I think is one of its strengths and something a lot of us need to think about doing because prehoc, you cannot phenotype patients for everything you may be interested in once you sequence them. So I think this brings up a lot of challenges with consent and how we work with patients. So we brought this patient back in and we did a history focused on this phenotype, which, of course, we didn't have because we started out studying cardiovascular disease. She had some history that was a little bit peculiar here, which you can see. But more importantly, we did more phenotyping in her, and what we found was that she had about 100 times the upper limit of normal metabolites in her urine and in her blood, suggesting that, in fact, and she also had, we scanned her, and she had multiple small infarcts, which are not dissimilar from the infarcts that the babies have who have the metabolic crisis. So really, what we can show here is that by the standard drill, yes, mutations in this gene caused the disease we thought we were studying. But the phenotype is much broader than we ever thought it would be, and we could only ascertain that by studying people in a hypothesis generating mode that is taking people starting from the variants and working toward the phenotype. So really, I think we should consider, and I'd love to work with people and share ideas about how we can do this, which is to take patients who are broadly consented, generate the genomic data first, sift them for the variants, make a hypothesis, and then do the clinical research. And we can only do that if we have well-consented patients. And currently, again, our presumptions about what we think diseases are really limit to the things we can discover, and hypothesis-generating research can potentially allow us to do, to study phenotypes that we don't even know we should be looking at. And I think, again, this is a way we could potentially work together. Before I do that, I would also like to say that this cohort is potentially available if other people have variants that they think are intriguing, that they don't have the ability to re-phenotype, contact us, and we can potentially collaborate with people to do that phenotyping. And it's actually been very, very surprising to me how willing our subjects are to be phenotype. And this has included things like multiple inpatient evaluations with IV glucose tolerance test to explore variants in genes that may relate to diabetes. So it's pretty surprising. Now, we can't do everything to everybody. There will be fatigue with the subjects, but doing this again in an open, iterative way can work. I think something that a number of us have been thinking about that we have to be very, very careful of is generating what David Goldstein has called just so stories. Because when you seek ones, you can always come up with a clever hypothesis that links your variant to your phenotype. And we have to be very careful not to overinterpret that because we have a huge Type 1 error problem here that we have to avoid and do the work in a rigorous fashion and be sure that these findings are valid. So I can stop there and take questions. Any questions? That's great, Aless. I'd just like to emphasize that I think if you look back over the last few decades, every time we've developed new technologies to look at our patients in different ways, the phenotype always broadens because we identify the phenotypes that we know how to identify and we don't identify the remainder of the phenotype. So I think this story will be repeated over and over again and biology and medicine will be the benefit for it. And I would just add yesterday the statement was made that you can't bring patients in to study them without telling them what the result is. And in fact, you can do that. And we can take patients who have been sequenced and go back to them and say, you know, we have to find variants in people and we need to study them further to see if they mean anything. And we would propose to do the following clinical evaluation on you to see if a variant is meaningful or not. And that is a legitimate mode of research and it can be done and it doesn't have to violate CLIA and other regulations. David has another comment. Did your 60-year-old woman have any Sibs? She does, yes. And we're working on contacting them. I think it's a small family. I think it's just only one or two if I remember correctly. So the odds are not high that we'll find anything there. But yes, we're doing that. And we do, again, have the ability with this cohort to go in and do segregation and then recruit relatives of these people who we sequenced to do segregation and that sort of thing. So Lester, question over here? Yes. Yes. So it's interesting that you found, you know, as you say, some of you weren't even expecting to have this condition. Have you considered, I imagine you have, going into already sequenced databases and finding people who are homozygous for this variant? Yeah. And, you know, many of them can't be re-phenotyped, but many of them can. And so could you do that? Yeah, we could. We've started that. We've looked in the 1,000 Genomes data set, and there were no convincing homozygotes or compound het's there that one could identify. But I'm sure there will be others. And a great place to look would be in that whole lexome sequencing place. Yes. Recognizing that many of them can't be re-phenotyped. But there is phenotyping on infarcts and other things. And so you might find out some interesting stuff. Right. But again, I think it would be, it would make all of our cohorts a lot more powerful if we could begin to incorporate some of this consent concept into them so we can all be doing this, because we'll learn a lot more. Oh, great question. I don't think it is, but that's always possible. Always possible.