 So first of all, as opposed to what Dick Winchelbaum said, I actually did follow the academic model and got everything done much, much too late, and did not have an opportunity to have a prolonged discussion with the panel at all. I did respond or react to Huckum directly about some of his material. There was primarily specific questions I wanted to ask. And now that I've been listening to the session here and participating today, as well as listening to Huckum's suspicions, I'd like to just go through a couple particular points and then have the John and Jeff and Cynthia respond. So I wasn't quite sure what was meant by different paths for phenotyping, but I think what I understand about that now is that just that there's adult phenotypes to pediatric phenotypes and that these are being done in different places by people of expertise, either in pediatric or adult phenotypes. I really don't see this as an obstacle at all. I think it's a strength in general. Pediatric research has been underrepresented in genomics, and so I think it's perfectly fine for there to be particular phenotypes that are enriched in pediatric patients and are not seen in adults, and that special attention will be put to studying them. I mean, this project, as I understand it from listening to Huckum, is that it is primarily a discovery project. This is not an implementation project yet. And so discovering genotype and phenotype correlations in pediatric patients makes a lot of sense to me. I had a question about the informed consents, and I know we discussed that a little earlier today. I had a particular question which has to do with what happens to pediatric data once the subjects become adults. I know there are some recommendations and some thoughts that keeping this data without re-consenting these individuals as adults is a problem. I see that re-consent process of tens of thousands of people as being a huge problem, and I'd like the panel and the people participating in this meeting to perhaps comment on that issue. I have some thoughts about it. I'd like to hear what people have to say. The next issue is, okay, is the TPMT genotype imputation. Has any of this been linked down to patient response? What fraction of these tens of thousands of patients have been genotype have actually been exposed to this drug, and do we have follow-up data and do we know what happened to them? I mean, you know, I think the impact from Mary Relling's work and other people's work on TPMT genotyping to guide a six-fire guanine or is a fire-cream therapy is well published and understood. I was wondering whether in this particular project we have any such information. I thought that the emphasis on the CNVs was also very welcome. I think they have been, they're obviously very important in clinical medicine and it's great to have them more involved in this discovery project. I just wanted to mention, it's something I mentioned to Hakon, that the DGV is a valuable resource, but there is ESCA, which has now become what we call the ICCG International Consortium for Clinical Genomics, and there's a major attempt made through a grant that's been funded by the NHGRI to try to get much more structural variant and copy number variant information linked to phenotype into ClinVar, into the public domain. And I think that having the eMERGE pediatric component work together with the ESCA or ICCG and to try to get as much information into the public domain as possible would be a great benefit for everybody. This is a small point about the concordance frame between sequencing and imputed haplotypes and it's not a big deal. It has to do with whether they're common variants or not. It's a rare variant that people are having more problems with imputing and of course that makes perfect sense, but for the more common SNFs, common variants, the imputation is far higher concordance than this and so I think that this is not really a major issue at this point. So I think to summarize my reaction is that I think many of the things that Hakon described as being obstacles I actually think are strengths and that's a good thing. I would like some comment made about the removal of pediatric data once these individuals have reached adulthood. I'd like to hear more discussion about going beyond just simply how well you are doing at imputing genotypes for TPMT, but instead is it linked to the EMR? What do you know about outcomes, phenotypes? How many people have been exposed to the drug? And I'll stop right there and ask the other members of the panel to comment. So why don't we'll pass the baton to Jeff Bodkin, who can perhaps as well as his job as moderator, moderate the answers to the questions that Bob has posed.