 So, I'm making a presentation on behalf of the training and career development team that includes Tina Gatkin, Gatlin, Heather Junkin, and myself. And Bob and Gail were co-chairing this session, which was held on the 10th and 11th of April. So, the content of this presentation will be briefly the purpose. The next two bullets, the overview and examples of genomic science and genomic medicine training programs was an effort to sort of educate the participants of the working group to the things that were going on at the director's level and at some of the selected institutions, institutes and centers at NIH that have sort of shared goals. And then the last bullet, we'll talk about, you know, really what actually happened during the meeting. So, for those of you who were here last May, last February, there was a small teleconference that included some of the members of the council to sort of start the discussion of how to align our training programs with the strategic plan. And that resulted in my presentation in May of 2012, sort of an overview of our current training programs, in which we had asked the program directors to give us information about their goals and if they were not aligned with the strategic plan, how they planned to do that, we asked them about some of the elements of their training programs. And we had also asked them to go out and survey a couple of their graduates to get some idea of what things they could do to improve their programs and as expected, most of them had very good reviews. But the next step for that was then to dig down deeper to give us more information about, you know, if we were going to move more toward the right, how would we do that because the programs that we now have are ones that were started at the very beginning of the genome program. And so they're basically at the left-hand side of the density plot. So the purpose of this workshop was to review the training and career development programs and to make recommendations on how to align the programs with the strategic plan. So just to sort of give you some background in terms of reference, we always speak in a lot of alphabets and acronyms here. So we talk about the F awards. Those are individual fellowships. And really there are four varieties that I've only shown three here. The F30 is an individual MD, PhD fellowship program. That was one of the things that came out of the director's working group on the biomedical workforce and that was to have all of NIH participate in that. Up until now, we did not. So we joined this and starting in FY14, all institutes will have to participate in the F30 as well as the regular F31. We have always participated in the F31 diversity program. The F32 is your garden variety individual post-doc fellowship. And there's also an F33, which is a senior fellowship that's very rarely used, but it is used by established investigators who want to go on sabbatical and use this as sort of some support while they are on sabbatical. Then we have the K programs are the career development award programs. The K01 and the K25 are mentored career development awards. The K22 and the K25 are what we call the two-phase career transition, which includes an advanced post-doc up to two years and a faculty transition up to three years. For NHGRI's purpose, a total of five years. And then we have the T32s, which are the traditional institutional training grants. So to sort of orient our working group, we had invited representatives from the office of the director and several institutes to talk to us about their initiatives. The three initiatives out of the NIH director's office included the workforce report on the biomedical workforce, where there is a concern that it is taking too long for someone to become a PhD. It's taking them too long to become an independent investigator. And the stipends are so small that these four things will now make becoming a scientist not attractive. So they are working on ways to improve that. The other directive out of the director's office is the diversity workforce. And that one has to do with some of the things that Richard talked to you about, and that is sort of incorporating a lot of these scientists into the peer review process. There are also two RFAs out dealing with mentorship and building infrastructure. Then Michelle Dunn from the Cancer Institute is leading the big data to knowledge effort in the training area. And she talked about the efforts there in which we will have a workshop in late July in order to talk about what are the training needs in those areas. So that is planned. And Mark is very much involved with that initiative as well as I am. One of the things that's always a question is when you're talking about what should NHGRI be doing, the other question is what are the other institutes are doing. We find it always difficult to go to report or any of the other NIH databases to really find out what other institutes are doing. So we thought we could help this effort by inviting some of the staff there to talk to us about their efforts. So the person from the Heart, Lung and Blood Institute was only talked about their undergraduate science institute in biostatistics. And that's a program that has been running for about eight or nine years. There are about six institutions that participate in that. And the success of that is that 60 or 70% of the graduates from that training program will matriculate into a graduate program, a PhD program in biostatistics or related area. The National Institutes of General Medical Sciences, as all of you know, it's the training program at the NIH. And we wanted them to talk about their efforts in bioinformatics and computational biology. And they also have a postdoc training program for clinicians in six clinical areas where they give them training in areas like medical genetics, trauma. But the one we were interested in was the one having to do with medical genetics because the idea was that this is something we could be using, looking toward our training in genomic medicine. They have, and IGMS has 11 postdoc programs, they have nine postdoc programs and 11 graduate programs in bioinformatics and computational biology. The National Cancer Institute, they also have programs in medical genetics, bioinformatics and biostatistics, and they have 18 programs in that. The Library of Medicine was very interesting in that they have 14 university-based programs in informatics and computational biology interfacing with healthcare, public health, and translation research. And the interesting thing about their programs is that they're very agnostic as to where their people land, and a fifth of their graduates are in industry. So we also thought it would be good to get some idea of our training program in genomic medicine looks like, and what one in genomic sciences look like. So we invited Mike Banky, who was one of our first program directors for T32, to talk about his program. And very broadly, he said that the program interfaces mathematics with genetics and genomics, that most of their graduates are, they major in one discipline and have sufficient foundation in a complementary discipline. The students matriculate one semester early, they come in the summer so that they get a head start for the semester. And he found that if there is a student who may not be totally prepared in an area like statistics or math, they will have tutors for them, or they may take courses to sort of get them ready for the regular semester. There are program enhancements like there's always a mock study section sort of so that the kids get a sense of what a review actually looks like. There are attempts to give them courses or information about writing scientific papers, about how you do interviews, about how you do presentations, and there is a lot of interaction on a social level so that the group forms sort of a cohort. About 90% of their trainees remain in research and we've always been able to provide additional funding for the training programs for things like tutoring or for recruiting and areas like that. And Mike, he was very strong in saying that training has to be broad and deep and that sort of goes against what the NIH model wants to be now by training, getting them out sooner. He also thinks that one measure of success is looking at how long it takes someone to, once they finish their undergraduate, to get a job. And I think his main point there is that most of the people who finish in his program, they don't linger in postdoc programs, most of them usually go right to a tenure track position because of the area that they're in. Then we had Jeff Vance from the University of Miami who is just starting a master's degree in genomic medicine for MDs. This is a four-year program that runs concurrently with a medical curriculum. The curriculum includes didactic courses. They also have labs, online modules, and small group discussions with the faculty. There are benefits to the program and that the students are now, they become pioneers. They're some of the first ones who will be graduating in genomics. They realize that genomic medicine is integrated into all of medicine and they also feel that having this master's degree will give them a leg up and getting a good residency. There are some serious challenges with this program in that it's very time consuming because the student is taking medical school courses as well as the master's courses. And also there's a cost. For some reason, the way the program is now set up, out-of-state students can get grants in-state students have to pay and it's like $8,000 for the course. So that is a negative in terms of trying to get people to sign on. Then we come to the real meat of the program and Eric got us off to a good start by talking about some of the challenges. Then we will look at the recommendations, the overarching principles, and recurring themes. So Eric identified four challenges for NHGRI in this arena. One is the pace of genomic advances and as he put it, depending on a lot of factors, the pace of genomic research can be a slow walk or speed walking or sprint and we don't know what factors will control the pace. So that is a challenge. We also target different audiences. We're looking at people all the way from undergraduate to established investigators. We're also looking at basic sciences as well as clinicians, so many audiences. And perhaps one of the real serious issues is the reduction in the appropriation dollars. So if we want to start new activities, where is the money coming from when our budget may remain flat or even decrease? And then the need to identify NHGRI's niche within the strategic plan. We always say that the strategic plan was meant for all of genomics and not just for NHGRI. And so within this big ball, we then have to decide what part of it is really unique to NHGRI. So to sort of prime the discussion, the staff here generated some questions and they were meant just to sort of start the discussion. It was not meant to confine the discussion but to give them a sense of some of the things that we were interested in, such as what knowledge and skills are needed, what are the most effective ways and environments to integrate genomic science with genomic medicine, how best to leverage training in the clinical areas because there are other institutes that are training in some of the areas that we are interested in and so we don't want to duplicate what they are doing yet we do have something unique to offer. What are the most effective ways to enhance training in the foundational sciences, and here we're talking about informatics, computational biology, the quantitative sciences for basic and clinical trainees and researchers. And what percentage of the extramural budget should be allotted for training and career development? Unfortunately, we didn't get to the last one but I think we did get the information that we needed to move forward. So the recommendations for career development, expand the K01 to cross-train clinicians and genomicists in genomic medicine. What this means is that as our announcement stands now, it is for individuals who have degrees in math, science, chemistry, physics, engineering and commuter sciences to train in genomics. So now we want to expand this so that anyone who has a degree in biomedical research or who has an MD degree can now come in for the K01 award. And there we're also saying that we should sort of not aggressively recruit this first group of individuals, continue to support the K01s and the K25s and the K99s, and to support individuals interested in technology development because the skill set that is needed for both genomic science and genomic medicine will always need these kinds of people applying their skills to the problems. And then to support individual K awards initially. And this came about because there was some discussion as to whether we should have institutional K awards. And the problem there is that these awards would be very, very expensive. And since this is a new area, it might be better to support individuals. And then once we've had some sense of what a training program should look like, we might want to consider an institutional training grant. And also by having individual Ks, this would allow these individuals to be spread out into a lot of institutions where if you have an institutional K, you'll only be able to make one or two of those and those would be clumped in one or two places. So the other thing we were very interested in is what constitute a good training program. And we mostly focused on genomic medicine because I think we have a fairly good idea as to how to train people in genomic sciences. But genomic medicine is an area that is really new to us. And this is the one that we were looking for lots of advice. So a defined general approach that is applicable across diseases. So one of the things they were saying is that someone can identify, say, a disease that has had a very problem that's been very difficult to solve. Well, they can work on that problem if they can demonstrate that the solution will be applicable to other diseases. They also should have a defined curriculum and training plan, an environment that is rich in genomic research. And many of the working group members felt that this was not a problem because there are a lot of institutions now that are heavily invested in genomics and clinical areas of genomics. So this should not be an issue. Access to critical resources such as raw data, clinical data, health records, just unmanipulated data that they could actually work with because this is the real world. And then dual mentoring and complementary disciplines. In terms of training, again, expand the training to encompass genomic medicine. But again, we still need people who are versed in the foundational sciences like informatics, computational biology in those areas, quantitative sciences, support individual and institutional training, allow flexibility in course requirements. And several people made the comment that comparing entering students now to those who entered maybe 10 years ago. Many of these kids now have a very strong background in computer science and the quantitative sciences. So when you start defining what is required, you have to look at the students that are coming in, and that should be left to the local level. To strengthen the foundation in genomic medicine by supporting training in the data sciences and LC, and limit training in genomic medicine to PhDs, MDs, or their equivalents. And this recommendation was made because they were saying, first, genomic medicine is a new discipline, and we don't really know what a PhD would look like. But you can take someone who has a PhD or an MD or an equivalent and then add on to that the genomic medicine training. And I think they gave, there are a couple of institutes now that have this kind of a training program. One of them is an IGMS that has a postdoc, it's an institutional postdoc training program, and child health also has one. And with both of these, they are geared to clinicians, but the advice here was to open it up to MDs and PhDs. The other thing is that there should be two paths for training, one, genomic medicine as a researcher, and then as a master clinical genomicist. And this would be someone who has foundational training in genomic medicine, but would have primarily clinical responsibilities. So that this person could be a teacher or a trainer to people who are coming up. Didactic training is essential, that LC training is also required, should be required. And that the training program should vary between two or three years depending on the background of the individual. And they said that four trainees were considered a critical mass for a viable program. And it should be noted that I know with IGMS their programs, the number of trainees varies somewhere between four and six. So in terms of the clinicians that they should enter during their fellowship year, they should devote 75% or more time to research and devote up to 25% time to maintain clinical credentials, but that their time in the clinic should be devoted to applying genomics in the clinic. We also support short courses that are not part of the K, F, and T's, but we should continue to support those especially because we've used those courses to disseminate knowledge to a larger audience. We have quite a few courses at Cold Spring Harbor and other places to pursue new methods of disseminating information, and here this refers to online courses, but the issue there is how do you evaluate those and how do you keep them up to date? And then collaborate with professional societies to conduct courses in genomic medicine. There were some overarching principles that applied both to training and career development, fund more training and career development, and the goals of all of these programs should be to expand the base of knowledge in genomic medicine, continue to support the foundational sciences, and to develop leaders. Again, the issue of training broadly and deeply, and training in behavioral and health sciences is also important. Even a professional network of trainees and K awardees was brought up. One of the things there is that many of the people in the audience that had experience with annual meetings of trainees coming together and sort of forming a cohort of people who would in future years sort of become members of a group, and then again to support the diversity action plan. Some of the recurring themes were increased diversity, use of a variety of metrics to measure success, and some of these included to recognize people who are PIs on subprojects, to recognize that a successful person could be someone who is teaching in a university that does not have a research intensive environment. Even people who are in industry, another measure would be like what is the impact of the publication, or has that person or their research changed the way medicine or research is done. Another theme was educating practice and physicians and collaborate with other units to leverage NHGRI support. So in looking at how we match up with the other institutes in terms of support, the people who came, the institute reps who came to the group provided us with some information and if you were just looking at the FTs and Ks, GM commits about 9.5% of its extramural dollars to these three activity codes, NCI about 3%, NLM 26%, a lot of their money goes into training. We commit 2.4% and the average for NIH is 3.6%. So the next slide tells you what this buys us. So for our KL1s and K22s, only for fiscal year 12, we had three awards. The average cost of that was about $150,000. It was 0.1% of the extramural budget. For the K22s and the K99s, we made seven awards. That was 0.2% of our budget and the average cost was like $80,000. For the T32s, we had 13, it was 2% of our extramural budget and the average cost was about $600,000. For our Fs, we made 12 awards. That was about 0.1% of our budget and the average cost is about $50,000. So that is just sort of a brief review of what we do. And I might add that some of these numbers included five LC grants. I think there were three Fs and two KL1s. But it does not include the LC training that is embedded in the SEERS. So the other thing is that this workshop only dealt with the non-LC part of NHGRI training. As Pamela said, the LC group has a parallel discussion going on with their community. And they will bring their information back to this group at a later time. So before I open it up to the council, I'd like to ask Gail and Bob if they have any comments. Gail? I think that was an excellent summary. Thank you. I think the committee was trying to balance the importance of clinical people in genomic medicine and also basic science and not losing focus on the importance of developing analytic people. Also being flexible about different mechanisms that can be used in what are efficient mechanisms such as K awards, but also possibly joining in with existing T32 programs for medical genetics, for example. Bob? No, I agree that was a great summary. I just emphasized the point that there was enthusiasm for expanding the efforts of training into genomic medicine at that distinct best doctoral level or beyond where you could take a strong foundation in genomic medicine with clinical expertise or strong clinical expertise. Okay. Heather? But I think you covered the highlight. Okay. Heather and Tina, do you have any comments? And I'll just remind you that the full report is attached to the agenda. So the council? Jim? I was just wondering what we know about demand on the side of those with MD and PhD equivalents in the sense that one of the things our field has struggled with has been at the slightly junior level of, say, at the medical school side, demand has not been high. Demand for what? For training, for further training as, you know, genomic medicine specialists. Yeah. And especially, you know, the demand among potential trainees. Are they going to see this as a unique and special opportunity that really gives them a leg up? Or are we, you know, do we run the risk of creating something that won't be? So I don't know if Bob or Gail want to comment. I mean, I can tell you in my going around the places, that's one of the things I've had folks at various universities advocate for, the belief that they have these clinical trainees who actually are getting interested but see no vehicle to find funding to allow them to take the time off of their clinical pursuits to learn more, to train. Bob or Gail, do you want to add to that from your experience in Seattle? Yeah, I think there's no question that that's true, that it's very difficult to identify short term funding. And particularly these people sometimes come up on somewhat short, shortish notice, you know, not years of planning. And we've definitely lost some opportunities. I mean, I don't have lots of experience with the clinic these days, but one area that I've certainly seen interested in is trying to develop diagnostics without an obvious vehicle system. Didi and then Tony? Yeah. So first I'd like to say I like the recommendations and it seems like it was a very good workshop and I'm sorry that I missed it. Yes, we invited quite a few council members who couldn't make it. But I do like the recommendations and then I was curious, just because I don't know that much about it, but you mentioned the Jeff Vance program in Miami, which is a master's in genomic medicine, and then the recommendation is only allow PhDs and MDs, but is my understanding that that master's is something that MDs get in addition to their medical training? Okay. And how many people are involved in that training program? Do you know? In the first year, I don't remember the numbers. I think it's like five. Yeah. I remember five. Five. Three people remember five. And I think some people joined up and then sort of bailed out because it was just too much time. It's a very early thing. Yeah. Just that might be useful as a gauge to the question on how much interest or demand is there. And I really want to emphasize the Miami part. I mean, if they're in their first year, I would regard it as an experiment and I think we want to keep in touch with Jeff and others just to see what this thing looks like over the first couple of years. And then just kind of going back to what we were talking about earlier in this open session on the report on the incidental findings and physicians and laboratories and so on. I mean, it just seems like there should be a great need for this kind of training. So maybe, I don't know how we better advertise that to make sure that people know that this is available. Well, when it becomes available, I mean, it's not available right now. No, I know. But yes. If we move forward, there's no question. If we move forward on it because I would think there would be great demand. And I'll just also mention that at Arizona State University, we have a department or program on biomedical informatics, which is geared also for this kind of thing for training and ideas to have that kind of training for physicians. Tony? I guess coming back to my former question, so maybe on a more macro level like other professional groups like physicians or dentists or veterinarians who look at markets and decide that there's some kind of gap about where the trends are going and how many they need to train. Is that ever done at NIH to look at the number of jobs that are out there in biotechnology, industry, academia to decide how, if there's a gap. I mean, I understand the working group is made up of academics and physicians are trying to understand where within the field you want to put emphasis. But I'm concerned that we might be overtraining the number of people that there are jobs for. I mean, I can take a pass at saying that one of the working groups of the NIH's advisory committee as the director that yielded report last June was on the workforce plan in Shirley Tillman. And I think it was Sally Rocky, I think was the co-chair. Somebody from NIH was a co-chair. I mean, they looked at that, but that was more in line with how many PhD, it was more at the graduate school level, how many people are you putting in the pipeline, what should be the trajectory of those individuals, what are all the things we could do with more, with the current number or more or less doctorates. I think this speaks to a slightly different issue. And I think that may be one of the reasons why this group that came together recommended dealing not with early stage but later stage. To the extent we need more people, in this case, recommended genomic medicine, what could we be doing to grab some of the physicians and turn them into physician-scientists, genomicists. So your question is, do we really need them? And I don't, I'm not aware of any specific study. I think it's anecdotal hearing from where people are trying to recruit. We certainly would love to have some more genomically trained physicians working in NHGRI, probably all different parts of the institute. And if Bob wants to comment as a chair, you might have experienced from Tufts what's going on. But it seems to be the impression that places are hungry for such individuals. But I don't know of a study, I'm speaking anecdotally. So I would just second that, Eric. And probably in multiple clinical departments I'm betting, right, Bob? Not just your department. Right. I think, I mean, what we're seeing is that our graduates are in high demand. And Mike Benke had some impressive numbers of graduates from his program. And even someone like Mike, if I recall, right, was saying, I mean, Mike's mostly training more basic oriented folks. And he was saying, wow, but in thinking about it, certainly people who are medically trained as well would be particularly in high demand. Oh, absolutely. I can tell you, it's very hard to identify these people when you have open positions. They just are very hard to find. Yeah, I would echo that. We are searching for a medical geneticist, and that's very hard to recruit for. But the other item is the data and informatics committee, or work group that Joe and I were in. I also advocated for more training in informatics, because that's a perceived need in the community. I'm about to be going back to the physician, genomist, as I wish Les Beesinger was still here, because I could tell you that Les, when he became a branch chief in her intramural program, was involved in a couple of rounds of searches trying to find somebody for his branch. And with that exact target of physician scientists working on clinical problems in genomics, and they're just not out there. Very much interested to have them. I think it's fair to say that our training budget is very small, so I don't think we would overwhelm the community with the number of people that we support. A question on, as I said, on the larger level, does NIH give guidance on the number of training grants they should be funding in an environment that they're cutting back research? I mean, if they're not getting jobs in the academy, where are these people going to go eventually? So that's why I'm saying that some sort of market analysis should be done to understand how much you should be either increasing, decreasing, or keeping the same, the training budget. So I don't know the answer at the macro level, but I do know that as a follow-up, NIH, in fact, I even think they're bringing in some, or as a consultant, some economists to really look at this. But again, Tony, I think they're speaking to a slightly different issue. It's more about how many PhDs. I think what we are looking at coming out of this is much more of when you have PhDs, when you have clinicians, we want them a little more specialized in genomics. Is there the need? And it's what Bob and Gal were saying, and others have been saying. But it is relevant that you're putting more money into that specific area, and you're not cutting back on all the other training you're doing. So I can see that the training budget is growing when everything else is shrink. And that's the reason for my sort of more macro comment. I think that'll be what we will have to discuss as we keep having discussions around this, is if we're going to do some of this, and if we move forward on what some of the recommendations came from this workshop, do we do this by adding money, which means it has to come from signals, or by shifting money, and that we're going to grab with. Okay, Carlos and then Bob. I mean, this is kind of almost more a point to Tony, but I mean, I don't think we should be guiding the PhD training program or the postdoc training program by the total NIH budget, right? Just because that's flatlining doesn't mean that those individuals aren't getting a competitive advantage by being trained by NIH or whether the country shouldn't be training more individuals in that area. I mean, even if you look at growth and finance, right, they're hiring individuals who are at CERN, right? The people who are part of the Large Hadron Collider are some of the most coveted by the finance industry. So, you know, I actually think it's a mistake to kind of fall into this trap that, well, NIH is flatlining. Therefore, we don't need to be producing any more PhDs because the goal of the PhD should necessarily be to find a job in academia or something where you're going to be funded by the NIH. I didn't say that. I said a market analysis which looked at a wider picture of what was in industry in other areas because these people have to get jobs. And what I see is you end up with a lot of people on their third or fourth postdoc because there's no jobs out there. So, just, you know, think about that when you start increasing training budgets. Right. I mean, the other thing to say about it, which we need to talk about is, first of all, whatever we put in framing, it's a pebble in a quarry compared to when you add an NIGMS and NCI and NHLBI. So, I don't think we're not going to spend any curves even when we went all in with our training budget. But second of all, we are way... I mean, other than they came out of this, which I think we want to think about, I don't know what we want to do, but I want to think about it. It's just as a percentage of our total budget, we are way under compared to most of this other incident. Even if we went to the NIH average, we'd still be a small pebble in a large quarry. But still, should we be that far below the NIH average? And that's what I think we want... And that's what I think we want to talk about. Oh, wait, I had Bob next. It's a really interesting report. I'm really sorry I missed the meeting. I wanted to be there. I had two questions for you. One is I was struck by the comment about being grateful to some of the NIH institutes that came to your meeting and spoke there. Is there no general cross-NIH clearinghouse on who's training and what kind of training so that we actually know what the whole NIH portfolio is? Does that not exist? You mean the training portfolio? Yeah. Yeah, you can get that information, but if you want to say genomic medicine, you will get... the SARS will pick up genomic medicine in anything, and you may get a lot of things that really when you look at them, they're not genomic medicine. That's the problem. So the website of the search tool is not very... Exactly, exactly. That's the second time today that I've found that a cross-NIH website search tool was not functional. Is there a theme here? Sorry about that. Second question, and this may be a purely narrow view that I have, and I've been struck by the fact that in a lot of MD-PhD programs, the number of people who are getting PhDs in bioengineering as part of their MD-PhD is actually quite small compared to other PhDs, and in the bioengineering, the number that are doing bioinformatics or computer science is a tiny fraction of that. Now, maybe I have a very skewed view of that, but I'm wondering whether the MD-PhD programs are another place where we could really try to get people trained both in medicine and in computer science. I worry... I guess I worry a little bit about how effective and how deep you can be to take someone who's a clinician and try to turn them into a bioinformatician without having really rigorous, fundamental teaching and knowledge about information science. I really worry about that. So I would like to respond to your comment about the MD-PhD training and having somebody in computer sciences. We could very well go out and say that for the F-30 program, we're only going to be accepting applications from people who are interested in an MD in computer science or informatics or computational biology. We could do that. Other people on council think that would be a good idea. I mean, is that a good source of the people that you want trained with this idea in mind? What's at the MD-PhD? Yeah, I'm asking other people on council if they think that would be a good thing to do with F-30. Well, the good part about it is that it's an individual award, so it's not institutional. So, you know, if you can identify somebody who was interested in that training, that would be great. Go ahead. Yeah. Then you and then Howard. Yeah, go ahead. I would just say that some of our most successful graduates were engineer majors that then went to medical school and so not necessarily an MD-PhD, but they had that background. The problem has been medical schools accepting those non-traditional pre-med candidates. So that's where there's more trouble, I think. Some of the audience is going to be sub-specialists that want to learn more about genomics. And a lot of the way you described it was that. So not so much they'd be board-certified in genomic medicine, but rather they would have that extra training. And that speaks to NHGRI as the ultimate leverager in terms of institutes, because it's going to be a cardiologist wanting to do two years of genomic medicine, not a genomicist wanting to do two years of cardiology. So it seems like the one solution or one way forward is really partnering with these institutes to develop these programs, partly because your budget's limited and partly because that's where the talent is going to be. Now there are going to be some programs like we've talked about with informatics and medical genetics that are going to be solely in this institute. I think most of the need is to make oncologists better genomic medicine people and cardiologists and that sort of thing. And that really is the specialty IC venue. So like a lot of things we would do, there's no question that once we put some meat on the bones of this we would absolutely go around and try to convince other institutes to participate because you're right. And all likely their proposals likely would have a disease-specific element or component or maybe that would be the dominant feature. And it would be... Eric? Yeah, Bob. I mean that was one thing that got a lot of attention I think in the discussions of the committee was just what the role of NHGRI was in doing this as opposed to the categorical institutes then. And I think there was a lot of enthusiasm in trying to get the categorical institutes to focus on their particular diseases and apply genomics to them and train people there. But NHGRI's role, on the other hand, would be training people with more fundamental advances in genomic medicine. Yes, it makes sense to me. But I also think there's probably a little bit blurry between them. I mean, it's not a sharp line. Right. And I know... And obviously if you can get lots of people... I mean to leverage this across the institutes would be fantastic. Right. And we're not going to get anywhere unless we develop the framework for this and put our skin in the game. If I may, Eric? Please. Yeah, so the... I mean I think it totally makes sense to leverage the other institutes and try and get them to accelerate their interest in genomic medicine. I would be a little careful about being too strict and defining who could be funded by NHGRI in that it's a very flexible field right now and there are lots of different aspects and costs of it. So I wouldn't want to say, well, only if they're going to do informatics because we have a guy who right now is an internist medical geneticist who's taking two years to get a master's degree in outcomes research focused on genomic medicine. So you never know where the opportunities are and you wouldn't want to be overly restrictive. Okay. Any other comments about this workshop and this discussion? David? I think a lot of the comments had to do with the applications to medical genomics and I actually like the fact that the workforce also recognized the need to maintain foundational training. I think it's just important to remember that genomic science is also transforming basic research with challenges that involve large data and statistical analysis in ways that also raise training challenges and the institute should keep its eye on the application of genomics to basic training as well as to medical training because I think that the activities are transforming both activities and it's important to maintain a workforce that can take advantage of the data onslaught that's also coming from the innovations. I guess I'd like to say that that was very clear and I think it's clear in the report that we should do that and I think perhaps the report sort of overly talks about genomic medicine simply because that's an area where I know very little about in terms of training and since we had the people there, this would be the opportunity to get some sense from them what would constitute a good training program but it's not that we plan to abandon our other efforts. The committee was very clear on that David. There was strong enthusiasm for the programs that are out there and the strengths that they bring to it. This ad complimentary to that was the emphasis on not continuing to develop people in technology and genomics that was good. Okay thank you Betty. We should hit the