 Thank you for the kind invitation. So I've been asked to talk about translational biology, what is next. So I'm just going to tell you what, in my view, what may be next. I think each of us may have our own list. So I think to me, first, we need to define the role of the checkpoint inhibitors. I think this is a class of drugs that we are all excited about. Because unlike the targeted therapies that we have, these are drugs that could potentially cure patients with metastatic renal cell carcinoma. And I think we want to do that, both of the checkpoint inhibitors alone initially, and then perhaps later on in combination. And along with that, it's going to be imperative to know what is the patient population that is most likely to respond to these inhibitors. Is PD1 or PDL1 expression going to be a good biomarker? So I think that's a very important question for us. The second question that I thought was worthwhile investing in is trying to understand the determinants of tumor aggressiveness. I think all of us in the room that take care of patients with metastatic renal cancer may have some patients that do not require treatment for an extended period of time. And then we have other patients that, despite our best efforts, die in six months. And that's something that we don't understand. And this is despite the efforts trying to map the cancer genome. Something else that was alluded to earlier today is late recurrences. So we may have seen patients coming with metastatic disease. Often it's just one or two or three metastases, but sometimes more extensive metastatic disease a decade or 15 years after the diagnosis of their primary. What is the underlying biology for that? And what determines tropism? What determines whether metastases go to the brain or whether they go to the bone? It's a problem that a memory-slang-catering is being tackled by John Massagé. But I think we need models that are more relevant to what happens in patients. I also think it's important to understand the determinants of treatment resistance. And it was alluded earlier today that the sonitinib and seraphinib and passopinib and axetinib, those are drugs that are not targeting the tumor cell. They are acting by targeting the endothelial cell on the periscide. So elucidating the mechanism of resistance is going to be a lot harder than elucidating the mechanism of resistance of each GFR inhibitors or BCRA of all kinds of inhibitors and so on. And I think this is a far more important question than understanding the determinants of sensitivity. I think there, I'm not sure that knowing what patients are most sensitive to the drugs that we have available today, it's going to help us particularly because those patients are already being exposed to those treatments. I think we need to understand and exploit mutation heterogeneity. We're all familiar with Girlinger's study in the New England Journal of Medicine. I think that can be both good news and bad news. It can be good news because it can help us identify the truncal mutations that we should be focusing on. I think it is important to try to isolate biologically distinct entities in renal cancer and develop a classification of renal cancer that goes beyond histology. Right now we are classifying tumors based on their appearance under the microscope. And as we've seen in the TCGA, the TCGA aimed to study clear cell venous carcinomas and yet they got a handful of translocation carcinomas. So a classification that's purely based on morphological features and immunochistochemical stains is not a pure classification. In addition, some data that I'm gonna show you will hopefully illustrate the point that there are different types of clear cell venous carcinoma with different biology and different outcomes. And I think we need to exploit recent discoveries about the molecular genetics for drug development. So obviously, and we heard from Billy Kim, how he's trying to do that, looking for synthetic lethality with kinases, which is a class of drug for which we have targets today. So these are the questions that come to mind with respect to what is next. And what I'm gonna tell you in the next few slides is what I in my laboratory have tried to do to get to personalize or precision medicine. And I'm gonna be focusing on the last two points that I have outlined here. So I think we are all familiar with this and this burst of drugs since 2005 and at the root of this progress, we realize that there is the discovery back in 1993 by Linenhan colleagues of mutations in the VHL gene. And the VHL gene is inactivated either by mutation or epigenetically in approximately 90% of tumors. And in fact, Sato et al recently showed that another protein that is part of the VHL complex is also inactivating a subset of tumors. So this pathway may be inactivated 95% of clear cell dinosaur carcinomas. And this is really the foundation for the development of these five drugs. And there is a second pathway which links with the HIF pathway and that's the pathway governed by mammalian target of rapamyzing and we know we have two other drugs. But I think the point I'm trying to make with this slide is simply that therapeutic advances are gonna come from a better understanding of the molecular genetics in the biology of renal cancer. So our goal, my goal at UT Southwestern has been to establish a research program that would ultimately deliver personalized medicine. And I think to get there, where do we need to start? I believe we need to start with the genes. We need to understand what are the genetic drivers that are mutated in renal cancer that are affecting the process of transformation. And knowing the genes is not enough. One needs to understand the pathways that are regulated by these mutations. And knowing the pathways is obviously not enough. Having, realizing what proteins in the cell are being affected by these mutations may give us targets that we can then use for drug discovery. And we need better preclinical models. I think it's fitting to try to have an effort along the lines of what Chris Wood mentioned this morning with respect to trying to develop better tumor graph models. And we're gonna need biomarkers so that we can tell what patients' tumors have what molecular genetic alterations. And most commonly in practice, immunohistochemistry is used. I think it's appropriate to try to have immunohistochemical biomarkers. And then finally, to bring these to the patients, we're gonna need clinical trials. So this was the program that I started to develop at UT Southwestern when I arrived. We've made some progress. And I'm gonna be talking to you today about our efforts with respect to the molecular genetics and how that establishes a foundation for a molecular genetic classification. So what do we know about the molecular genetics of renal cancer? Well, so I'm gonna illustrate a few manuscripts that I think are most important. So this is a manuscript coming from the Sanger Institute describing the identification of set D2 mutations in renal cancer. And we heard about that from Ian. Here we have a second manuscript also from the Sanger Institute showing mutations in polybromone. And I will be talking some more about that. So a manuscript from the Beijing Genome Institute where they report the identification of approximately 15 or 20 genes that they found to be mutated in clear salvinous carcinoma, one of which has been validated in other studies and that's BAP1. This is a manuscript from our own lab showing that BAP1 is mutated in 15% of clear salvinous carcinomas and that BAP1 loss defines a new class of renal cancer. And then we have a couple of recent manuscripts within the last few months. This is the manuscript coming from the TCGA and this is another manuscript coming from Japan. So I think these manuscripts overall provide a very good foundation for an understanding of the molecular genetics of renal cancer. So I'm gonna focus on two of them. So this is the first one. So polybromone. Polybromone is a two-heat tumor supressor gene. Like VHL, both copies of the gene are inactivated in clear salvinous carcinomas. These mutations result in truncation of the protein and that is very often loss of the protein in the tumor. And besides the truncating mutations, there are other mutations. So overall, PRM1, I would say it's inactivated in approximately 50% of clear salvinous carcinomas. PRM1 encodes BAP180. This is a component of a nucleosome remodeling complex. And it's thought to regulate the inaccessibility. And we heard from Ian about this. So I'm not gonna go into this, but essentially these nucleosome remodellers will regulate where the DNA attaches to these histone octomers and regulate the inaccessibility. So this is the manuscript from our own group. So we found that BAP1, like PRM1 or VHL, is also a two-heat tumor supressor gene and it's inactivated in 15% of clear salvinous carcinomas. BAP1 mutant tumors tend to be of high form and grade and they are associated with MTORC1 activation. And this is different than the PRM1 mutant tumors. Mutations in BAP1 and PRM1 tend to be mutually exclusive, not always, but they are underrepresented, the simultaneous mutations in tumors. BAP1 and PRM1 mutant tumors have different gene expression and different biology. BAP1 mutations frequently cause loss of the protein and this allowed us to develop an immunohistochemistry test which in fact performs almost, or perhaps one could say better than mutation analysis. So you can see the two-prositive rate of this immunohistochemistry test is 25 out of 27, whereas for mutations it's 24 out of 27 and the true negative rate is very comparable. So we believe this establishes the foundation for a molecular genetic of arvinous carcinoma in which in our series we had 55% of the tumors that had loss of PRM1, 15% of the tumors that had loss of BAP1, including 3% of the tumors that had loss of both. And then there were 30% of the tumors that were unaccounted for. And interestingly, as was alluded to previously, the VHL gene is very close to the BAP1 and PRM1 genes. In fact, you have loss of this region and chromosome 3P in the vast majority of sporadic clear cell arvinous carcinomas. So the 3P loss causes the loss of one copy of VHL, CD2, BAP1, and PRM1, leaving cells very vulnerable to the loss of the second allele. Interestingly, as I have mentioned earlier, there is a negative correlation between BAP1 and PRM1. So if there are mutations in PRM1, the likelihood of having a mutation in BAP1 is reduced by 70%. In contrast, there appears to be a positive correlation between mutations in PRM1 and CD2. And this data is based on a meta-analysis that we did using our own data and data from the TCGA and others. So this has led us to propose the following model for venous carcinoma development. So we believe, and there is some evidence to support this, that clear cell venous carcinoma begins with an intragenic mutation of VHL. And this is followed by loss of 3P, which results in a loss of a copy of these four tumor supercell genes. And then tumors may be divided into tumors that have mutations in PRM1. These are tumors that are likely to be of low grade. And tumors that have mutations in BAP1 and these tumors tend to be of high grade. And when we look at survival, and this is just simply looking at our cohort of patients, we find that patients with BAP1 deficient tumors tend to have a significantly higher probability of death compared to patients with tumors that are mutated for PBRM1. And we made the same observation looking at the TCGA, and James Cie, who is also in the audience, has a similar data from memory as long catering. So we've gone ahead and entered into what I think has been a very fruitful collaboration with Mayo Clinic and with Rich Joseph, who I believe is also in the audience. And we've examined using our immunohistochemistry test a cohort of 1,400 patients for whom they had, as you can see here, outcome data as far as 20 years. So these are patients that are presented with localized or local regional disease. And we simply did the immunohistochemistry, gave them back the data. And as you can see, these curves separate very nicely with a hazard ratio that is very similar to what we saw before. So in other words, it appears that patients that have tumors that are deficient for BAP1 have a significantly worse renal-specific survival that patients that have, that are whose tumors are competent for BAP1. And I'm not gonna show you this data, but this is in a manuscript that is in press. BAP1 status predicts independently of UISS and SCOR and in low-sign patients. This is data that Rich is putting together and it's actually very nice. And further extends the observations that we have previously made, showing that patients with tumors that are well-typed for BAP1 and PBRM1 do very well. Patients with tumors deficient for PBRM1 seem to do just as well. Patients with BAP1 deficient tumors have lower or have decreased survival. And then patients with tumors that are mutated for BAP1 and PBRM1 have the worst outcomes. I should point out that here the instrument, the immunohistochemistry test was essential. Oops. And the reason it was essential was that in some instances one can find mutations in BAP1 and PBRM1 in two separate populations of the tumor. But by immunohistochemistry we could tell that the mutations had occurred in the same compartment. Okay, so I'm just gonna finish up with just making you guys aware that it was previously described that BAP1 mutations occurred in the germline and we found that BAP1 mutations are associated with familial and renal cell carcinoma. So you can see here this is a study that we did in collaboration with Marston Linehen. We examined approximately 85 families with familial renal cell carcinoma that was unexplained and we found one family that had a mutation in BAP1 that cosegregated with a phenotype. This is not the only family that has been described. It is a second family that was recently described from a group in Europe where BAP1 mutations also predisposed to renal cell carcinoma. And I'm just gonna go to the acknowledgment slide. So I'd like to thank for this work. So the genomic analysis was done in collaboration with scientists at Lumina. I've alluded to the work done by Rich Joseph, Alex Parker and their group at Mayo Clinic. The collaboration that we had looking at the familial renal cell carcinoma was with Marston Linehen, Laura Smith, as well as Charis Eng and the Cleveland Clinic. Kim was involved in the project, facilitating access to TCGA data and looking at outcomes. And I'd like to thank the people in my laboratory who did the work. The work on the molecular genetics was spearheaded by Samuel Peñalopis, which is an instructor in the laboratory in the work on the molecular genetics of familial renal cancer was spearheaded by Megan Farley. And we worked very closely with our colleagues in urology, as well as radiation oncology, pathology and radiology. Thank you for your attention.