 So, first of all, I also would like to thank the organizers for the invitation to speak. This is an exciting meeting with, I think, really critical outcomes. I'd also like to say that I love to really like Mike's talk, but the last four slides or five slides I could have used to summarize my talk. The conclusions are pretty much the same. And one thing that really resonated was that in the human genome we want lots of function mutations, big-time ones for every single gene, and that's one of the messages to come out of what I'm going to say. So I'm going to address two general questions. The first is, what is the value of Mendelian genomic research to medicine and society, because I think this isn't necessarily widely appreciated by many in the now expanded human genetics community, which were concerned with common diseases and genomics rather than Mendelian disorders or medicine. And secondly, the second question I looked at carefully at the achievements of the centers for Mendelian genomics, and are they making a substantial contribution? So first of all, Mendelian diseases with conditions, which I'll summarize MC, it's easier. MCs are individually uncommon, but collectively, they add to a huge burden of illness and disease and costs, 0.4% of live births, 8% of adults. So not all single gene diseases present in the first few years of life. There's a gradual expansion. Huntington disease would be a good example. About 25 million Americans and cost about 5 million per individual, per lifetime, for those who are born with such a defect. So this has a major impact on the overall health of Americans. One of the major messages I want to convey is that we're not even, I think there's a misimpression that we're sort of halfway to finding all the genes associated with Mendelian conditions. And this is far from the truth. First of all, there's 7,300 Mendelian conditions presently identified, but if you look at the bottom line, there's three, to these three genome centers, CMGs, there's 300 new phenotypes being presented to them per year by people who are not even focusing necessarily on phenotyping. And one of the major contributions of the CMGs is try to enhance the ability of physicians and geneticists to do the phenotyping. The second thing is that there are only 2,700 disease genes identified compared to the 20,000 genes we know are in the genome, and I'm going to talk more about that. Of the about 7,300 diseases identified, about half of them are explained and about half aren't explained. So there's a lot of work to be done here and let me elaborate on this. So about 17,000 genes of the total human genome remain as Mendelian candidates and I should have changed this, what this should say is that 10,000 of these genes we're totally agnostic about. We don't know what they do. You can make a guess and say that about 6,000 genes in humans might be embryonic lethal based on mouse knockouts. But if you have hypomorphic alleles in those genes, you might very well get a human phenotype. So it's extremely possible that this 10,000 plus this 6,000 can all result in human disease and in addition to that there's another 1,000 diseases which are mapped but the gene has not yet been identified. So that's where the 17,000 come from in contrast to the 2,700 genes that have been linked to a phenotype that explained about 3,500 Mendelian conditions. So the point there is that a single gene can be associated with more than one conditions and I'll come back to that later. Also there's no evidence whatsoever that I could find that the discovery of new phenotypes is becoming asymptotic. So one area I work in is retinal degenerations and this is from Steve Degger's website called Retinette and I actually thought this was going to be when I hadn't looked at it for about three years flattening out. In fact, if anything probably with the rise of next-gen sequencing, the number of identified genes in the concentration of precise phenotyping has increased, continues to increase about the same rate as it has been over the last 15 years. In addition, totally unexplored and here's where we're lacking biology is we don't know the contribution of non-coding genes very much to Mendelian disorders. So there's about the same number of non-coding genes as there are coding genes and these include long non-coding RNAs, long intergenic non-coding RNAs, short non-coding RNAs, micro RNAs. Maybe these are overestimates. There's a recent paper that I just saw this morning published in PLOS genetics suggesting that only rather than all of the genome producing products that influence biology, maybe only 8.2% of it does. So this is something for you to look at and consider. But nevertheless, I think we would all be surprised if a vast number of non-coding genes are not implicated in Mendelian disease. So this is another huge unknown. Here's about 20 Mendelian conditions for which the evidence is firmly associated with non-coding genes. This is in an article by Cilio Santonarakis in clinical genetics last year. So why does identifying the genes for these conditions matter to patients and their families? I can tell you, for those of you who are clinicians, you know, but if you don't, it matters hugely. It's everything to these families. If there's no diagnosis, there's no prognosis, okay? There's no best practice guidelines. There's no reproductive counseling. And most in particular, there's no available therapy. And one of the themes that's repeated when you consider Mendelian genetic diseases, they're the ultimate in personalized medicine. If you don't know the gene and often if you don't know the allele, you can't treat the patient properly. When there is a diagnosis, you can know about the natural history. You can counsel accurately, sometimes prevent complications. And this little image shows the dilated aorta that occurs in Marfan syndrome, which can be slowed, if not prevented by the administration of a drug, Los arts, and providing tailored therapy. So not only does the study of the Mendelian conditions help families tremendously and contribute to medicine in that way, but it also contributes to our understanding of complex diseases. I think a theme which Michael touched on. And if you haven't read this amazing paper that was in cell last year, set aside about three days and do it, it's a very dense and I think very fascinating paper. The question being asked was, are common variants for complex diseases enriched within loci associated with Mendelian comorbidities? So they found that each Mendelian variant highlights a subset of genes that are also important for some common diseases with the same basic biology or organ affected and so on. Secondly, the conclusion was, and I'll show you some of this, each complex disease has a unique Mendelian allele architecture, a so-called non-degenerate code that identifies each illness by its associated Mendelian loci. So this is a tiny representation of the total dataset, but if you look on the lower right corner, single gene diseases that cause degeneration of the basal ganglia, the genes that are associated with those diseases are also identified by GWAS studies to be associated with a very high relative risk for Parkinsonism, PIX disease, Alzheimer's, and general frontotemporal dementia. So the Mendelian conditions are good signs to indicate where to go when you're looking, trying to sift amongst the genes the 100, let's say for diabetes that have been found in that study. I'm sure some guidance has been, can be found. So overall we observed that complex, these authors, complex disease genome-wide association signals were globally enriched about two-fold in Mendelian loci. So there are three, as Eric said, three Mendelian genomic centers, University of Washington, Yale, Hopkins, and Baylor, and their goal of them established about two years ago, and they've done a great deal of work, as I'll show you. I was tremendously impressed by what they've done. I hadn't followed this at all and spent some time searching. The goal is to overall identify and define the causes of all human monogenic diseases. That's the broadest goal, and everything flows from that. So first of all, the CMGs are an international research platform, not just confined to the USA. To do this, as I think Michael indicated too, I loved one of Michael's points, which was the necessity for collaboration has brought together the research community for genetically complex diseases in a way we couldn't imagine, and that's the same thing for Mendelian conditions. And so the countries in yellow are those outside North America. There are 568 investigators that are participating in 255 institutions. So this is a tremendous act of leadership on the part of the CMGs located in the United States. So what is the biomedical and clinical impact of the work that's been done by the CMGs? The progress to date, and this is information I got from the CMGs, is pretty amazing. We're talking about two years, okay? This is not the summary of all Mendelian studies over the past 30, this is two years. So 15,000 samples have been studied from 6,000 families, and 673 of them, the gene that was involved was known to be associated with that phenotype, but the clinician in the Senate hadn't recognized it. 760 novel Mendelian conditions were identified. So a clinician sends in a family and doesn't have a clue what's going on, and in 760 cases, the relevant gene was identified. So this fact indicates how important it is to have excellent clinical phenotyping, again a point that Michael Benke made about genetically complex diseases. So about 11,000 whole exome sequencing and 60 whole genome sequences were done, half of them were in the DbGaP database. 286 novel Mendelian conditioned genes were discovered and 229 known genes were identified. And the clinic, here's another important point, that there's an interplay between the genome centers and the clinic. So the clinical features, when you find a phenotype that you can't assign to any particular gene, and then you find it's involved, it's due to mutation in a known disease gene, but the phenotype has expanded. The clinical features, that occurred in about 139 cases. So this is called phenotypic expansion and that's gonna increase greatly. And you can see that the CMV's coined the word phenotropy, I had never heard of this word before, which means the spectrum of, it isn't even on the internet, spectrum of phenotypes caused by variants in a gene. So most, now I think there's a widespread misconception if you have a, what's that? Now, pleiotropy is the different manifestations. Okay, this is quite different phenotypes. Plyotropy is the pleiotropic effects, different effects on a body of mutations in a single gene. Let's have the debate afterwards, yeah. The Broad Center's not allowed to interrupt. So based on the fact that in many cases, so first of all, the first point is that there's a conception that once you find a mutation in a gene for a Mendelian condition, that that Mendelian condition is due only mutations in that gene. That is absolutely not true. And the converse is not true that one gene doesn't have one disease. So look at Collagen 2a1 presents with 16 different phenotypes and you know that Lamin has many different phenotypes associated with this. So the predicted novel genes for known Mendelian conditions is suggested to be about 3,000 based on the fact that when they sequence cases of particular phenotype, they still hadn't found the abnormal gene and significant fraction of them. So identifying Mendelian MC genes greatly enhances our understanding of human biology and pathophysiology. And this is evident from the publications from the CMGs, 98 papers including 60 new disease loci and genetic disorders in very excellent journals, Nature Science Cell and so on. And I'll just show you a few of them. So first of all, the number one journal in medicine as I think you all know is New England Journal, Impact Factor 55. There's two lovely papers in there from the center's CMGs. And then that led me to ask the question, how many papers overall on Mendelian diseases are have been published in the past couple of years in the New England Journal and Beth Finniser told me almost 5%. So this is, even though these are rare diseases, the biological and medical significance of these discoveries is highlighted by the fact they're published in this really outstanding journal. One of the most impressive papers I came across from the CMGs is this one, de novo mutations in histone modifying genes and congenital heart disease, but this paper is a lot more than that. So it's the most frequent birth defect congenital heart disease is many cases of sporadic suggesting you might be a role for de novo mutations. So 362 severe cases were analyzed in trios, whole exome sequencing, and there were two major findings, enrichment of mutations in proteins that modulate H3K4 methylation and de novo point mutations and several hundreds of genes that together give rise to 10% of severe congenital heart disease. This is a group of, this is a clinical disorder about which, prior to this study, we knew compared to this denominator, knew virtually nothing. The second example of deep biology that's informed by some of the work of the CMGs is this one, human clip one mutations alter tRNA biogenesis affecting both peripheral and central nervous system function. Clip one is a kinase required for tRNA splicing. It's present in almost every tissue. And why the mutations in this gene should lead only to neurological diseases unclear. And this is a question that arises with many different Mendelian condition genes. The genes widely expressed, but only one tissue is affected. And so there's a lot of biology there to be elucidated. A key part of the whole activity of the CMGs is to improve phenotyping so that clinicians worldwide who are not necessarily working at the Hopkins Genetics Clinic or University of Washington or Baylor can contribute their cases in a knowledgeable way. So to do that, phenode database was established a new web-based tool for collection storage and analysis of phenotypic features. And the main focus of this database is to allow rapid and efficient entry of families or cohorts provide unique identifiers for the phenotype. So the way it's structured encourages you to be precise in your entries. Clinical features are based on OMIM clinical synopsis and it's searchable. I think this is actually a probably an understated major contribution whose real value has only become apparent over the next five years. So identifying the genes for MCs is of major importance to the development of treatment for common diseases as well as for the Mendelian conditions themselves. So Mendelian genes identify drug targets applicable to the general population. So there's about 350 proteins that are linked to a human gene and specifically targeted by current therapies. That's kind of amazing. You think there's only 350 given the diversity of diseases. About 40% of these proteins encoded gene underline a Mendelian condition. That's a very surprising figure whereas about a little more than half of the proteins identified by current therapeutics are found in GWAS signals. I'm sure that's number is going to increase as well the number associated with Mendelian conditions. So some of the examples of novel drug investigations that are being pursued as a result of Mendelian genetic discoveries are mutations in the NAV 1.7 sodium channel. There's a duplication and no deletion. Channel, channel, which is given rise to which causes loss of pain and so now there are small molecule drugs which inhibit this channel's activity which are in clinical trials. Similarly, this channel, potassium channel is affected in barter syndrome which is characterized amongst other things by low blood pressure. So again, clinical trials are underway for small molecules that inhibit this channel to see whether they're effective in lowering blood pressure. I think everybody's familiar with PCSK9 protease and the family of small molecules that's being tested to see if they lower cholesterol just as the deficiency of this gene lowers cholesterol. The rexin receptor mutants cause narcolepsy and so therefore if you could knock down the activity of this you would have an effective sleeping pill. Another clinical set of clinical trials is underway. Mutations in these genes lead to high bone mass and therefore might be useful treatments for osteoporosis and we think we're all familiar with the roles that EPP and gamma secretase play in the biogenesis of Alzheimer's disease and so if we could inhibit the abundance or activity of these two proteins we may be able to improve the outcome with Alzheimer's. Okay, gene therapy is now real and for monogenic diseases, as I'll show you in a minute, some dramatic, one dramatic example, tremendously effective. So of all the gene therapy trials currently underway, 1966 of them, the majority are in cancer, 1270, but 8.9, the second most commonly area for gene therapy studies is monogenic diseases, 178 trials are underway around the world at the present time. So I'll give you one dramatic example for those of you who are pediatricians in the audience and have ever looked after a patient with this horrible disease, metachromatic leukodystrophy. Here's what this child who died shortly thereafter is spastic and has no really cognitive function whatsoever at the age of five years. So Luigi Naldini and his team in Milano have done gene therapy on three patients who were pre-symptomatic and it looks like that the disease progress has been completely arrested. So it's gonna be take a long time to see whether it's been stopped dead, but this is really compared to the alternative, a wonderful outcome. So as I said before, Mendelian conditions, the management of them is the epitome of personalized medicine. You have to know the knowledge of the individual's genome and affected genes in order to take care of them properly. And so this has given rise to the orphan drug industry. There's about 200 companies now conducting orphan drug clinical trials. It's a $50 billion industry growing rate of 25% a year. And there's been some amazing successes and this is a lack of coffee at six in the morning, but Kelly Deco and Luma-Caftor, which are the two of the drugs now being used for the treatment of cystic fibrosis, are dramatic examples of the success of the orphan drug program. And I don't know if you've noticed that the Kelly Deco, which was originally only for one particular allele of cystic fibrosis has now been proved for another eight. So it's nine altogether. And the combination of Kelly Deco and Luma-Caftor looks like it's effective in Delta F508 patients as well, but the jury's still out on that. But you know, these are fantastic successes and I'm sure they're going to be replicated by more molecules that will have effects on different alleles of the cystic fibrosis and in other diseases. And these are two of the websites that you can get this information from. So if we're going to support the mock, then one thing that concerned me in thinking about this talk was that there's a huge focus and investment in the mouse knockout project, phase one, which is underway in 18 centers around the world to cut their teeth, knocking out 5,000 genes, requiring a commitment of 250 million from multiple funders, including the NHGRI, an additional amount of money which I wasn't able to elucidate for embryo phenotyping and about 3 million for nuclease gene modification, technical development. So all of the biologists I think realize the critical importance of the mouse knockout project. And it's going to continue on in a second phase when they hope to tackle the remaining 15,000 genes. Well, Sydney Brenner a few years ago in nature, I thought made a dramatically accurate statement which is humans are now the model organism. We know how to study them so effectively to not be making that your first thought when you have a question to ask about medicine and sometimes biology is probably not the way to go. And so if we're thinking about the human equivalent of the mouse knockout project of course is the characterizing, identifying all the genes associated with a Mendelian condition and identifying all of the Mendelian conditions. And this is, I think one thing I, Michael slides about what would happen if you didn't continue to fund the the genome wide association studies and so on could be applies exactly, I should have seen your slides before I talked to the funding of the CMGs. I think the outcome would be I think at least unfortunate both for Mendelian condition research and for complex diseases as I've suggested. Another point to make about the mouse knockout project is that many Mendelian human conditions have no mouse equivalent. And this is a particularly frequent story for immunological diseases where our immune system is really quite evolved and different in many ways from that of mouse. And this is one paper and if you look up Jean-Laurel Casanova's work, you can see many examples of this, of diseases which are strictly human. And I'm sure that this is true for many other areas of biology as well not just the immune system, that's just where you get an initial dramatic presentation and therefore we learn about it. So some general principles likely to be exposed by a large collection of Mendelian disease genes as the relationship of, as I said, of genes and variants to phenotypes. So as collagen 2A1 indicates, one mutations in one gene can give rise to many phenotypes. The phenotypes that are associated with mutations in a gene can be expanded so the basic core range of phenotypic features is there but as you see more and more patients, the phenotypic features associated with mutations in that gene can be extended. The elucidation of, if you think of Charlie Boone and Brenda Zandru's yeast knockout project which has been in science a couple of times where they're using synthetic lethal analysis to identify genes that are critical and cluster them into biologically functional groups, the Mendelian conditions are the human equivalent of that project and of course it'll be many years before we're close to that depth of analysis. The relationship of Mendelian genes and variants to those contributing risk for complex traits I've talked about extensively, that paper in the cell by last year and a library of drug targets will continue to expand and be generated from Mendelian study and analysis of drugs for Mendelian conditions and I think it would be kind of surprising if many of these drugs or relatives of them are not applicable to genetically complex diseases. So why have the CMGs rather than, and this is something I dwelled on for some time and read the literature of what they've been doing, rather than just distribute this money widely in the research community. Well, first of all to do this with the degree of depth and sophistication and analysis that is required, the concentration of expertise in the CMGs facilitates this. So it's cost effective rigorous and a productive access to the best technology for experienced and naive investigators. So one thing I didn't emphasize in that heart study for example is that there were four clinical groups in Germany or centers, pathologist clinicians who contribute many of the samples in addition to clinicians at Yale. And I could give you dozens of more examples but the authorship of these major Mendelian gene discoveries is increasing greatly because you need the clinicians who are expert to contribute well characterized phenotypes where you're likely to get the same hit or subset of hits. Other advantages are that the CMGs are immersed in broader issues beyond finding the gene for a particular patient, rather the issues of Mendelian genomics. For example, what is the contribution of MC genes to genetically complex diseases? And these are issues that are applicable to other projects, technology development, another example. I think a last important feature is that the CMGs are agnostic to clinical area. So it's not a lab or a team that's like say Ed Stone's at the University of Iowa who does wonderful work in identifying retinal generation genes. That's looking into the street light to some extent and in contrast the CMGs are totally agnostic to this if the patients are well characterized and genetics look clean, then they will be studied by the centers from Mendelian genomics. So a summary could also be a restatement of the opportunities which will I think continue to expand. And so in other words, I think the CMGs should continue to do what they're doing and they will continue to do it better. So first of all, CMGs have made relatively inexpensive high throughput gene discoveries for these conditions available worldwide. You can't have every center with the degree of sophistication and sequencing and analysis of it that too widely dispersed. Enormous amounts of information about the biological function of each gene is provided by each Mendelian condition solved. Changing the thinking about the extent of pleotropy and genetic heterogeneity. So that relates to the number of conditions associated with each gene and the number of genes associated with one condition. Enable the CMGs have enabled diagnostic and predictive testing for hundreds of Mendelian conditions that were undiagnosable previously. And these patients are fed in by, you know, often very sophisticated clinical groups. So this isn't the low-lying fruit. Added hundreds of starting points for the development and testing of targeted therapies. As I mentioned before, only about 300 proteins are targeted by current therapies. And lastly, catalyzed the discovery of genes underlying Mendelian conditions. Hundreds of new phenotypes and novel genes for Mendelian conditions have been delineated. That's gonna be, I think, increased in the thousands. Thousands, not hundreds of novel genes will be continued to be discovered. New biological mechanisms. Look at the tRNA paper and cell associated with neurodegeneration. New biological mechanisms for Mendelian conditions. The histone methylation paper on congenital heart disease. Foster development of statistical framework for assessing the causality of variants for Mendelian conditions. And this is the point at which your average, even quite sophisticated clinical research group, just couldn't, I think, couldn't begin to approach the expertise that's found in these centers. And equip investigators in the wider human genetic community with tools and skills to interpret, and in many cases, complete their own analyses. And that's where the richness of the involvement of the other countries and the investigators there comes in. Thank you very much. We just have time for a couple of quick questions. If Dr. Rodin could come up and get your slides ready. No questions from the brode. Yeah. Okay, Sharon Plon, Baylor College of Medicine, not the brode. So one quick comment and one question. At some point you said something about the solved cases where the physician had no idea what was going on. As a physician, I just wanna clarify that those were the physicians who realized that these were patients with unique syndromes and did the work to submit them. No, that's not, I didn't mean to deem, that's a very important point. No, but the point I was gonna make is I think one of the things moving forward, if one wants to a large scale, is we continue, we need to continue to make it as straightforward as possible for physicians to identify these patients, be aware of the CMG and be able to submit them. Cause I think for scale, that's gonna be important. My other question was just, there was no comment about regulatory mutations. And you were talking about non-coding RNAs, but I think one very large class are, which it's difficult without scale, is to really identify regulatory mutations that cause Mendelian disease. I agree and that's where whole genome sequencing rather than whole exome sequencing, I hope, I think we all hope, will be informative. Yes. So I thought the analogy of humans with mice and yeast as a kind of knockout system was very apt. One thing to that is kind of, I know this is a really hard question to answer, but do you have in your mind, I was trying to look at the report, a kind of rate of solving per year per cent or was that just so such a dangerous number to talk about that you don't want to talk about it? There were, I had 96 slides at 10 o'clock last night and got them down to this number. I took those slides out and probably Mike Bamshad can answer that question better, but I recall it's about 40 per cent of families the gene has found. Mike, do you have another comment to add about the rate? But it's the right case? You had a specific graph and I can't remember what it said about the rate of solving. It depends upon the specific question that you're asking, but if you ask about the solve rate for phenotypes that we've studied, it's about 55%. Okay, but that's a kind of percentage of the families coming in. How many family, how many per year? So that's not a percentage of families. That's actually a percentage of the Mendelian conditions. So diagnostic rate is per family, which is about 25%. So now if you translate that into a rate per year of discovery, given the inputs, what does that look like? Yeah, that's a good question, but we've only been around for a couple years. So this actually, for this report, was our first sensors wide attempt to estimate a success rate. So we don't have a per year rate yet. Just one point on one. If I could just comment, if you simply ask how many new Mendelian disease genes have been discovered and divide that by the total number of samples, it would be about 40 samples to discover a new Mendelian trait locus. Okay, so the, all right. I think an important comparator here is that at clinical labs, excellent ones, they've forgotten the percent, but none of mine, I'll take too long to explain. Any other questions? We have to move on if there's one more. Yeah. I'm just curious about individuals with deleterious mutations with no or minimal phenotypes. Do you think there's a role in investigating all these patients, do you think Genomes? You know, I've been practicing medicine a long time. I mean, it's hard to think of many genes in which you have a knockout with no phenotype, and albuminemia. Your albumin gene doesn't make anything. It's a completely benign trait, except maybe if you get into a state of septic shock, but I think that's gonna be, I don't know, it's a good question we don't know, but it's, there's surprisingly few that I know about that are benign. And don't forget, it depends on the environment that one finds themselves in. I mean, some of the pharmacogenomic genetic diseases are only revealed by the exposure to the drug. I'm sorry, we do have to move on to stay on track. Please, thanks.