 Thanks. Well, it's a pleasure to be here. And that by Mark was just actually a bit of an inside joke. The mark also needs no introduction. The So Dan showed this slide earlier about our pathway and a slide that he has shown and I have shown a lot. And you know, one of our perspectives on this and how we make this a reality is of course the foundations of biomedical research and commitments to information technology and this iterative process of using the healthcare system, electronic health record, all the data we passively collect as a tool to engineer and adapt the system over time. And so because of this, and many of you I know are familiar with a lot of the work Vanderbilt has done, I just want to set up as a foundation that, you know, we use electronic health record to both feed research and implementation and a lot of the work that we've done to implement actually involved, you know, using our data to show that, you know, these experiences actually matter for our patients and it's led some of the iterations. So on the discovery side we have BioView which is de-identified that's not used for clinical care. I just want to set that up. And implementation side where I'm going to spend most of this time talking about is predict program which led to our ignite project. We did not imbibe as much alcohol before as Indiana and did not come up with this, you know, as creative a name as ingenious. And that's also led to some implementation activities around eMERGE as well. And that's all in the CLIA environment data in the electronic medical record. And as others have done it's been part of it primarily as a quality improvement initiative as opposed to research initiative. And then we look at what happens. And the discovery side BioView is formed from a de-identified resource of all the electronic health record data and other tools we bring in there. And that's about 235,000 individuals that we do research on. And, you know, a lot of that has been on disease associations but it's been a good tool for research into pharmacogenetic effects. One of the first we did as we were designing and launching predict was replicating this well-known association with Clopidogrel. And we've talked a lot about it so I'll just show you, you know, the clinical trial data versus, you know, our data and you see essentially the same effect size. And honestly that was really important for our physicians because they were saying at the time I didn't necessarily, you know, I don't necessarily believe the research data. I want to see it in, you know, kind of real world data when they take 15 other medicines whether it matters. And so, you know, we were showing, well, you know, this is what happens. Now that was maybe a different era. That was 2010. But, you know, it was an important aspect of it. And so we've done a lot of other discovery efforts here. It has not yet led to, interestingly, a new discovery of a new pharmacogenetic effect that we have turned around and implemented. But it has led to, for instance, a few revisions of some things we did. For instance, we first moved from alerting providers for both poor metabolizers and intermediate metabolizers, really based largely on the results that we had for Clopidogrel. So talking about predict and our implementation program, that's from our first flyer that we handed to patients. One of the things we decided to do first was define, you know, kind of who is at high risk if we were to implement a preemptive program, which was our desire. So we looked at 53,000 patients who have routine care at Vanderbilt. And we asked how many received one of, at the time, 57 medications with an FDA story for a pharmacogenetic-based prescribing. And we found that over 65% of those patients would have received at least one of those medications. You can see some of those individuals received quite a few, you know, with a significant number of people receiving 10 or more medications that had potential pharmacogenetic indications. And if you look at just six drug adverse events that have high morbidity or mortality, that equates to about 383 events over five years for those 53,000 people, or about 12 to 18 events for an average PCP. So this was one of those things we said, okay, this gives us some evidence to suggest that these events will be relatively common. Interestingly, it's skewed towards some of the more frequent medicines ended up being skewed to some of the more frequent medications with more serious adverse events like Clopidobrill. We did some surveys of providers. This was early on into the implementation of our program, but it was actually after it had been implemented. These are 121 people who had seen a patient with pharmacogenetic test results in their chart at some point, not with any direct education efforts or anything outreach from the program. And you can see by the top bar there that, you know, the vast majority of the physicians interestingly believed that genetic profiles would influence a person's response to drug therapy, that most of them believed specifically for Clopidobrill and also Warfarin that these would make a difference. And the factors influencing whether or not they would consider ordering a pharmacogenetic test was driven most strongly by their belief in the evidence of efficacy, which we found very reassuring. The fact that it was an institutional priority, for instance, is ranked towards the bottom of this list. Being prompted by an alert in the EHR itself was actually at the bottom of the list, but the interesting thing is people don't remember to order it without having prompted by an alert. So they won't order it because of it, but they will not order it without it. So when we set up Predict, one of the other things we did is we set up a review process. It starts with evidence-based review and guidance for professional society. CPIC was not really in existence when we started this. And so now a lot of that has been supplanted by CPIC guidelines and we are able to follow those and we do follow those and are involved in some of those now. And then the replication component that I mentioned before. Some of the replication was actually needed to give real-world models to what we were going to provide suggestions for, for instance, with Warfarin. And then we have a special division of our Pharmacy and Therapeutics Committee that would review these and then present these to the rest of the P&T committee before being implemented. This is what our intervention looks like when you open up a patient's chart. We have a homegrown EHR currently for another six months and then we'll be moving to EPIC. In this face page of our patient summary, we put the drug gene interactions between the allergies and adverse drug reactions and medication section. We don't necessarily expect people to look at these and act on and remember to look at these drug genome interactions when they prescribe, but it gets people used to seeing these kinds of results. And that's one of the reasons we put it up here. It also shows you what we had implemented, or do you have implemented? Clopidogrel, Simba Statenbier, Purine's, Tachrylemus on there, and I feel like we're missing, did I say Warfarin? Warfarin, it's five of those. In addition to that, we, of course, have clinical decision support on both our inpatient and outpatient environments for all of these. And that was really something we considered a requirement before we released them into the patient's chart. Those five interventions were released over the period of a couple different years as the evidence was developed and the decision support was developed. In certain cases, we had to enhance the underlying EHR to support, for instance, the calculations needed to calculate Warfarin dose. And so it took a couple of years to actually get that into the system. This is what it looks like for Clopidogrel. You're able to change your prescription with a single click. Dan showed this result before. This looks at the first about 10,000 patients that we had tested and how many had actionable results. So Todd talked about about 25% of their patients would have an actionable result for a given DGI. And Chris, I think you echoed similar numbers from the 1200 patients project. This cumulative result of 91% would be if you encountered that drug, essentially if you encountered all five of those drug genome interactions or sort of drugs in a patient who was genotyped, you would do something different than normal for 91% of them. And if you look at each individual drug, with the exception of the thyropurin example, each drug does have about a quarter of the patients who have an actionable variant on either an intermediate or a poor metabolizer status or it could be ultra rapid depending on your drug. Since we multiplexed this testing, it gave us, and a lot of these tested individuals were tested preemptively, it gave us the ability to look at our ability to reuse the data over time. And as I said, you know, we rolled out these five interventions over the period of a couple different years. And so the original people were just tested for Clopidogrel, and that was the only live test for probably the first year or so. And so the first 4,700 people that were tested were tested and had an immediate indication for one drug. And about 53% of those individuals at that point were essentially tested preemptively. And then as we unveiled more drug genome interactions in patients accrued more drug exposures over time, 4 years into the project we had used the data over 14,000 times on the 9,600 people we had tested, although about 6,000 of those people that were tested initially had no indication for being tested. And so not everyone who was tested did we use their data, but certainly some individuals we used the data multiple times. And so there's a cost savings here, even in a small number of drugs, to test multiple potential indications upfront in this analogy. When you look at what providers do, this looks at whether our providers followed a recommendation for Clopidogrel. And again, all these individuals will see decision support. A number of these individuals are tested prospectively. Some are tested as they walked into the cath lab, so the test results would come back after they potentially received their stent and had already been prescribed Clopidogrel. But you can see that providers responded in really a dose-dependent fashion based on the severity of the variants. So those that were poor metabolizers for CYP2C19, about 58% of them eventually got switched to alternative therapy. If you remove the contraindications for Prasagrel, which was the primary drug during most of this time, about 70% of the people were switched that were poor metabolizers. And then about half that number that were intermediate metabolizers, 33% were switched to alternative therapies. And of course, the alternative therapies also cost more. You have to recontact individuals in some cases to switch as well. But you can see providers are sort of making a real calculation on whether or not they want to switch individuals based on a number of factors, some of which is, of course, I think, perceived risk of the adverse event. As newer drugs have come out, you know, certainly there's a prevalent argument that the newer drugs are better. We should just switch to using those. And, you know, Clopidogrel maybe is being replaced. So we are, as Ignite, looking at a network, whether or not we're seeing this kind of effect, and what are the opportunities to tailor therapy based on genetics are. And so these are some initial results, looking at, in this case, anti-platelets. So you can see Clopidogrel is still a predominant medication used at three of our Ignite sites, Vanderbilt, the VA, and Aurora still continue to use Clopidogrel predominantly over Prasagrel and Ticagrel or the same thing is seen for warfarin. You can see a more dramatic, there is an increased prescription of the alternative agents in the case of warfarin, especially at the VA. And the VA, as many of you know, has a strongly driven formulary, and their formulary decision was new starts could use alternative agents off the bat. And they feel like that's ultimately more effective and is cost efficient for them. So, but outside of that Aurora and Vanderbilt, we still see a lot of use for warfarin. And I want to give another example of one of our adopter sites and their adoption and some of the influences of this. So Sanford is in the Dakotas and they have adopted a lot of the CPIC prescribing recommendations and had been going on for a couple of years. And this is actually showing the number of genotype for Clopidogrel coming through the cath lab. But one of the things that's really interesting is you see this inflection point here and this really results, this inflection point occurred essentially after Lori presented our data from Ignite on the outcomes from, with major adverse cardiovascular events following those who are switched to all genotype driven therapies versus not and showing that those events essentially decreased alternative agents in a significant way. And so this individual director of the cath lab became a supporter and their adoption has increased. We're also looking at cost effectiveness and this is work by Josh Peterson. He has a website that they've put up there on the bottom there. It's in beta and have done some initial simulations looking at the incremental cost effectiveness ratio per quality here. And you can see that the cost is pretty low with some given assumptions for minor little frequencies, event rates, the changes in odd ratio of events based on, for instance, Clopidogrel versus Prasigrel or Ticagrel or, but you know, it's probably not necessarily cost efficient to genotype just for Simvastatin in this example. Now these are, you know, this is early work, this is given a lot of assumptions. You can go online and actually simulate your own scenarios if you want. But, you know, you can see some cases that might be a clear win. Other cases it may not be, you know, this I think points to other experiments which Josh and his team are doing, looking at multiplex testing and that may be where some of the true wins are in terms of cost effectiveness. A few of our lessons learned, you know, implementation is about many things and you've heard about a lot of them and I just kind of pulled it down to a few but, you know, our laboratory test, even something simple like that we haven't talked a lot about is still a bit of a bleeding edge and we dealt with issues there and have dealt with issues transitioning that in a clear environment. You know, we've all dealt with a lot, recently a lot of EHR upheaval and changes and we're going through that ourselves now. That will probably we think settle down as many places have now adopted initial EHR systems but it is still a transition point and every implementation has been different and that's both environment and EHR specific and so we need to think about ways I think of going past that as Sandy and others have pointed out. We need to think about local provider buy-in in many ways and that's belief in clinical efficacy is very important. Ease of use and familiarity with the system are all very important factors there. We found that advice changes frequently and opportunities to reuse the data over time is frequent and that drives a need for surveillance and getting back to patients and those are points that Heidi and others brought up earlier as well. I want to give you one real world example from our samples and then I'll stop. So this is a 57-year-old that early on in the program who came in, who had diabetes, first incidence of angina and any no-none heart disease and she was cathed, received a stent, copitagal started, that stent closed off and closed off again and she was restented again and received more stents and by 11 months later by December she had been admitted nine times, had five interventions with nine stents placed and of course was at this time noted to be a poor metabolizer and was switched to Prasigryl. This highlights the need for these sorts of examples and potential power for some of the extreme cases. This is clearly an outlier but these cases do exist and is I think a powerful story of where there could be some cost benefit as well. So with that I'll end and there are many, many people who are part of this as well as with everyone else who has talked. So thank you for your time. Thanks for your unstenting efforts in this regard, Josh. Why don't you go ahead and sit down and we'll open up our discussion period.