 Okay. Go on. Okay. So we're going to change gears a little bit, and so we were talking about ENCODE and NACRI, we're thinking about where does ENCODE lead to. And so I'd like to present a concept clearance for demonstration projects for the genomics of gene regulation. And so the planning input for this concept clearance came from a workshop, Genomics of Gene Regulation, which was actually a planning meeting which was held in Gaithersburg in October of 2009, but also in the early planning of the future genomics meeting. We received input there. So I'm going to go over briefly the genomics of gene regulation meeting. It was about a year and a half ago, and the questions that were addressed at the meeting, how close can we get as a scientific community to take the goal of complete understanding of the rules of gene regulation? What more will we need to learn in order to get there, and then what genomic approaches to this challenge are both valuable and feasible over the next five to ten years? Some of the conclusions, and I'm not going to go through the whole list of conclusions, recommendations from the meeting, but I'm picking cherry-picking ones that I think are of what helped guide us in putting together this concept clearance hour. In terms of basic knowledge, we need to obtain more detailed information about the structure and in the longer-term dynamics of chromatin and how it relates to expression. Another recommendation is obtain additional validated examples of gene regulatory networks. They also talked about technology development. As Elise pointed out, we actually had a concept clearance in February of this year for the technology development for high-throughput functional genomics that Mike Payzen presented. And then finally, their recommendations for resources projects was to establish and carry out high-throughput functional tests for DNA and RNA elements that control gene expression, prioritizing intelligently, including perturbation methods, and in the longer-term genomic manipulations. With these planning meetings in mind, we set forth to make a genomics of gene regulations concept clearance, where the challenge is to identify the regulatory elements and to understand how they interact to drive gene expression. So this goes to the broader goal of, you know, you like to be able to read a gene DNA sequence and predict where and at what levels a gene is expressed. We think that ENCODE and MODENCODE were a start. They developed an initial catalog of functional elements. And we have to be very careful about what we mean by functional elements here. The catalog that ENCODE generated were sort of functional elements that could be defined by a biochemical assay. So does a histone mark, does it map to a particular region of the DNA? This does not mean that we've identified all the enhancers, all the transcript promoters, et cetera, of the genome. That relies on integration of the data. I think we're getting better, but I think this needs to be improved. So again, the goals, we'd like to demonstrate that genomic data and genomic technologies can be used to advance the understanding of genetic regulatory systems, and I think that the way to demonstrate this is you generate a model that is experimentally tested. And I'll confess something here that this slide set initially set a predictive model, and I talked to Jill before this meeting, and I removed that because it's sort of the connotations that it brings, and I don't mean to put Jill on the spot, but I think what we're talking about is developing models and then biologically validating these models. So the scope, what we expect these projects to be is, first of all, to select a well established model system, one where you know a lot about it. Take advantage of encoded, modern code data where feasible, or if not, take advantage of other data. The linked data would be an example, the epigenomics data. We would like to generate additional data on gene expression signatures of different cell states in the model system if necessary, again, use available data sets if possible, and also if you need to generate additional data on relevant functional elements. So you may have a model system where encode has a number of the transcription factor binding sites for this model system defined, but you may need to find a couple of other ones in order to sort of develop the first model. And then once you have the data, you want to generate a model about the biological role of the functional elements and how they interact to determine the gene expression signature. Again, the goal is to identify the functional elements, identify how they interact to determine a gene expression signature. And we anticipate these awards will involve experimental testing of the predictions about the biological role of these functional elements and their interactions. So for instance, if you predict there's an enhancer in front of a gene, you want to test that either by putting it into an enhancer assay or taking it and knocking it out. Similarly, if you propose this enhancer, interact with the promoter of the gene, you'd want to come up with some assays to test that. If you think that a transcription factor is particularly important in this enhancer, you'd like to take that transcription factor, knock it down in this model system, and then be able to assay the gene expression signature. Does the knockdown, does the perturbation give you the effect, the outcome that you anticipate? Finally, you'd like to see iterative improvements of the model based on experimental results, the idea that as you generate more data on what works and doesn't work in terms of the data validations, you'll be able to develop better modeling. And then finally, we'd like to think that, you know, by comparison of the experimental data and the models between the different participants, we'll be able to address questions like are there generalizable rules for gene regulatory systems? So the mechanism of support and proposed budgets, we are actually proposing a U01 cooperative agreement, but we anticipate that they'll participate in a GGR research network. And it's going to be sort of not a heavy handed network, but one essentially that involves sharing results and making sure the data is released as it's available. We anticipate setting aside $10 million in total cost per year for three years. And again, we're thinking this would support five to eight cooperative agreements. So somewhere between $1.5 to $2 million total cost per cooperative agreement. Additional activities, and here's where I'm probably going to get hung. We expect the data to be handled by the, at least some of the data to be handled by the ENCODE data analysis and coordination center. Because we think that a lot of this data will be ENCODE-like data. Other data can go to other repositories. But we didn't want to, we didn't really want to get into the hassle of setting up an independent data core, these projects to run independently. This is not a situation we're expecting all the groups funded under this network to generate a single product. We're asking them to explore what they can do with current technologies and current, and current infrastructure. We also anticipate that we, we also put out a program announcement for R01 awards for to fund additional analytical efforts. This would involve no set aside funds, but they would compete against regular NHGRI funds. And what we initially thought about is just to support methods to develop better models of gene regulatory systems. And any, any sort of technology development that's required for, for the informatic handling of this data. And so I think that's where I'll leave it to questions. So I'll take questions now. Brett. I don't, I don't think you're going to get hung. I think that people thought that was a good thing to do. Well, the cost is the question because we, we probably underestimated the cost given the discussion that we had earlier. Can you talk about how things that would fall into that third category here might relate to things that fell into the third category in the previous one? Again, I think, I think since, since these are more exploratory centers on how to do something, they're not so much, I think the third tier things for the encode project were to enhance the utility of the data to develop new methods of, of either, you know, generate new biological insights or correlating with disease studies or improved methods. In this case, we're actually, it's more of an informatics technology development that will help this type of research that, so that you can improve on either, either interpreting the data or, I mean, I haven't said the word visualization, so that you can improve on the visualization of the, of the models that come out to allow the scientists to sort of get, wrap their heads around it. So I would characterize it as the difference between really a, a, a community resource project and GGR, I'm not saying it's not a community resource project, it would have, you know, defined data release policies, but it's, it's more of an experiment and these groups will be, you know, will pushing their, their technology of, and their system. Ross? So, this is harkening back to our previous discussion about encode and I believe Claire, you asked if, if the aim of the next phase of encode was to finish completeness and, and personally I cringe every time I hear this term we're going to complete, I'm going to get a complete catalog, but, and that's, that background, that's why I really like this concept clearance a lot and here's something where you've got a defined goal, you can quantify it, you can see how close you come, if our ability to predict patterns of regulation is only 20% of what you want to be, we need to find that out and, and we can move, and probably we, the field can move in an orderly way towards a goal that is well defined and I, I think this is just fantastic, the, the, the concept. Now, there's no, no dollars for the back in this, right? There's no dollars for the back. Yeah, so that just, I mean, I guess, just to state the obvious that I would emphasize even more than the, the, the need for, for those funds, but I really like the overall idea. But again, I look at this as I don't want it down. Any activity in this meeting, if you would like to stand in line and for others joined, please find one, if not, I'll end the meeting. Maybe that's a statement on my presentation. Yes. Again, I don't want to, I, I, I, I run into this struggle about, I don't want to, there's so much, the decks are somewhat problematic of getting data producers to, to submit to it and I don't want, these, these are demonstration projects and, and, and I, I refer to it as a loose network. It's not like we're going to, it's not like in code where you're on the phone call with us more than you'd like to say. So, but I do want that, that's why I say I wanted to try to get them submitted to the, to the encode data analysis. Is it, is it that the deck has to have the wherewithal to handle the expectations? And, and if you, if you're, if you're asking me to do so much, no one who's honest is going to put in a proposal, you know, it really has to be workable. David? I agree. I thought this was a very nice set of concepts. I guess though the one thing I'm sitting here wondering about is, would you make any attempt to urge investigators to, if they were going to pick a model system, a regulatory network to, to focus in on, I'm sitting here thinking, gee, would be nice if they would pick one that was relevant, relevant to human phenotypes. And not only to define it like, here are the elements, but in the end, what you want to know is what happens when there's a single nucleotide variant in one of these elements? How does it perturb the behavior of that system, which PS is relevant to a certain set of diseases? So I do certainly like that concept. The problem that I have with it is, is that we're doing demonstration projects, and there are certain advantages of working with model organisms where there are much more, your ability to manipulate them is, is much greater. And so I imagine there would be a mix. Again, how do you mix five to eight awards? But I imagine some might be on model organisms and others might be sort of a very well defined, you know, human cell model system or mouse model system. Well, I guess, then I guess that one needs to be clear when you say model, whether you mean model organism or a model system. I need a model system. So it could be that, you know, you take, you know, I'll use one example I have in my mind right now, mouse C2C12 cells, you have been a low syrup, they become muscle. You might examine the regulatory networks that's in that. Now that doesn't have the disease relevance. But it may be that you're, you look at induction of interferon induction of gene regulatory networks. Howard. So I wasn't clear why you picked the U01 mechanism versus going with an R01 or whatever the next ability to expand it. It didn't seem to be a lot of things that were that cooperative. You said yourself they don't, you'd like them to play together, but they don't have to. I think they have to play together in the sense that we want them to be submitting data to encode. So we want to enforce the encode like data standards. I think we have to enforce the data release policies. And I think this is a situation of you want this to be a community where we can get them together periodically once a year and say, what have we learned? And if necessary, you know, we could if we learn something, we can transmit that to further workings. You know, I hate to use the word working groups, you know, we can set up working groups within the consortium, within the loose research network of how we're going to work. And so that's the reason why I think I picked that as opposed to just, you know, here's an R01, here's your money. And well, we might call you back every year, but we wouldn't have as much, I want to say control, we wouldn't have as much ways to incentivize them. It makes the encode coordinating center aspect even more important because it is, it is the hub of the wheel. It's the only place outside of those annual meetings where people will be interacting. No more discussion. Well, actually, this is needy here. I just have a couple of comments. I couldn't tell if it was a time to time in. Um, I, I really like this concept, the concepts in this clearance. And I would say that the budget is modest. It's the 10 million a year for three years. I, these kinds of experiments that I'm using the spell based assays, all the computation is needed and all of that are are very expensive. I don't really have to, the projects will have to be built very carefully. But I see it as something that'd be very exciting. You may want to encourage teams with the computational experimental and maybe medical people for some of those comments that were made about having something for relevance to disease. Thanks, Dee Dee. And I should ask if Rick is on the call. He was having travel problems. He might be in the air. Okay. Any other comments? If not, can I have a motion one way or the other? I move approval. Ever. I'm supposed to move. I move. I move for approval. But by the time I said it, I think it was a second. Any other discussion? If not, all in favor of approving anybody opposed? Okay, thank you.