 Thank you. Good morning. I'm Stephen Kingsmore, and my team is doing research of two types. Number one, research to develop new methods for rapidly diagnosing children with genetic diseases, and then second of all, doing research to assess the usefulness of genome sequencing to identify the root cause of illness in newborn babies. By doing this in two settings, one, they're both level four or regional neonatal intensive care units, where care is provided for the sickest of babies. And one is in San Diego at Rady Children's Hospital. The other is in Kansas City at Children's Mercy Hospital. This shows our study. Our study is designed to measure traditional measures of effectiveness of genetic testing, such as rates of diagnosis and time to diagnosis, but also less traditional measures, such as what parents and physicians think about genome information in this unusual setting of an intensive care unit and an acutely ill baby. We're about two years into these studies, and I'd like over the next couple of minutes to tell you two brief developments that we've made in the last year. We're going to do that in the context of a baby who was enrolled in our study on April 10, 2013. This is a baby who was acutely ill. This baby's mum had been picked up during pregnancy as likely having an abnormal pregnancy. She had an abnormal blood test, and as a result of that, she was seen at Children's Mercy Hospital and received maternal fetal medicine care. At that time, it was noted that her baby had congenital anomalies, and specifically, as shown on the left-hand side of the slide, an abnormality of the stomach wall, or the abdominal wall, called an umphaliceal. So the baby was delivered in our Children's Hospital and admitted to the neonatal intensive care unit for treatment. The baby was doing well in the intensive care unit, as most babies do, until about two months of age when he developed acute liver failure, and that's shown on the right. So he was jaundiced, and his liver was not functioning. The cause of this was not known, and acute liver failure in a newborn baby is something which is a very serious condition, and the baby's pediatrician, Dr. Petrikan, had counseled his mum that her baby was likely to die unless the cause of this illness was found. And so this begins a race against time where we need to make a diagnosis of a genetic disease in order to identify a potential treatment for this baby's condition. And so you'll see over the next slides a time clock running, as I describe the methods that we've developed for this. Now, the reason why making a genetic diagnosis is so difficult is that there are an awful lot of genetic diseases, in fact over 8,000 of them, and that number increases by about 20 per month. These are diseases that are caused by mutations at one place in the genome. If we add them all together they're very common, affecting about 4% of U.S. children. And they're the leading cause of death in infants that's children less than one year of age, and in the neonatal intensive care unit and the pediatric intensive care unit. So this is a serious healthcare problem. The solution that we and others have developed to address this problem is to test all genetic diseases instead of just the ones which are high on the list of possible diagnoses for a child to test all of them at once by decoding the entire human genome and by doing that as fast as possible. And so one of the key findings that our research reported this past year was that we're now able to do that in 26 hours. We published this about six weeks ago and we demonstrated both that it was feasible to decode a genome and get a diagnostic test result in 26 hours. And also we described methods of making this more sensitive and more scalable. These methods that we had described lacked sensitivity and also were only applicable to one baby at a time. And these newer methods are things that we think we can scale up and make broadly available. So let's take a look at baby 487. So the first step is for mum and dad to give consent for their genome and their baby's genome to be decoded. Next we get a small blood sample from mum, dad, and baby. And we transport that to our genome center, our laboratory, which is inside the hospital, either at Reddy Children's Hospital or Children's Mercy Hospital in Kansas City. And the time clock is running. We're about 60 minutes of a lapse time since we were notified about this baby. The next we isolate DNA from the blood samples, and that takes our time to about two hours. DNA is the code of life, the genome is all of the body's DNA molecules, and it looks a bit like chewing gum. We then get the DNA ready for decoding with a series of chemical steps, and now the time clock is at six hours. So I want to show you a video, at least I hope I can show you a video, about the size of the human genome. It looks like we have a very slow computer here. So maybe we won't show the video and I'll just describe it. So the human body has 37 trillion cells, which is a difficult number to understand. And then to each of those cells is packed two copies of the human genome code, each of which is 3.2 billion DNA letters. And to try to help you understand that, as some of the other speakers have told you, the DNA code is a four-letter language, and it's written in sentences and paragraphs, and we read it just the way we would read a human book. But the book, the genome book, is the equivalent of a manuscript 400 feet tall, and that information is packed into each of the 37 trillion cells in the human body. So that gives you an idea of the complexity of the issue in terms of making a genetic diagnosis by genome sequencing. So we load the information onto a genome sequencing instrument made by a company called Illumina, and we're still racing against the clock. So we had been at six hours, and now we have run the actual DNA material on a sequencing instrument, one shown in the lower right-hand corner, and we're at 24.5 hours, and we have managed to decode the entire human genome. The next step is to make sense of that information, and this sort of slide starts to try to help you understand what we mean by understanding a genome. So the genome, as I've said, is 3.2 billion letters, or nucleotides long, and the key information in a genome is called genes. There are about 20,000 of these, and each of these encodes a number of proteins. It's proteins that actually conduct life's activities, and so we're looking at for changes in the four-letter DNA code that can impact a protein's function. So here's how we do this. We use specialized computers to make sense of this. So as of 24.5 hours into the process, we had decoded infant 487's genome 40 times, or 120 billion letters of code. We assigned to each position in the genome a two-letter code, so there are two copies of the genome, one from mom, one from dad. We need to know what the letter is at each location from mom and dad's genome transferred to the baby. We then look for all of the letter changes in the baby's genome that differ from what we would call normal or the reference genome sequence, and typically we have about five million letter changes, so we still have a very big issue in terms of understanding the cause of this baby's illness. We're able to get rid of most of those changes though because they're common, they occur in more than one in a hundred people, and those type of changes couldn't cause a rare genetic disease such as liver failure in a newborn, so we can ignore those limiting our search to about a million DNA letter changes. Next we're looking for letter changes, as I mentioned, that can cause a genetic disease that's letter changes that can affect a protein's function. So here is the DNA code of one particular protein called hemoglobin, which is a constituent of red blood cells, so a red blood cell, a normal looking one, is shown in the left hand bottom side of the picture, and hemoglobin is one of the major components, protein components of red blood cells and assists in transporting oxygen from the lungs to the body's tissues. What's shown above is the start of the DNA code for hemoglobin beta, and if we change a single letter, an A to a T, at this position in hemoglobin beta, that is necessary and sufficient to cause the genetic disease, sickle cell anemia, and if you have two copies of this letter change, one from mom, one from dad, and you have sickle cell disease, then if the oxygen concentration in your body drops, your red blood cells take on this characteristic change in shape, which is associated with clinical symptoms. So we're looking in our baby's genome for letter changes of this type, and we're able to do that with computer programs, and in this particular baby, baby 487, there were about a thousand such changes, which is typical in a human genome. Next we take the baby's clinical features, what are the symptoms and signs of disease in this baby, and they're listed on the left hand side, and I've mentioned that the baby had unfalliceal and liver failure, and some of those other terms are things that are associated with liver failure in a baby, and we use a computer to match those clinical features to the 8,000 known genetic diseases and to rank order the possible diagnoses in our baby, and in this particular case, there were 341 potential genetic diseases that could cause liver failure, and this really helps us understand how difficult it has been traditionally to make a diagnosis of these type of genetic diseases in babies receiving intensive care in a manner that's timely enough to change the treatment of the baby. So when we apply the 341 disease list to the 913 changes left in the genome, we're down to two, and that gave us a diagnosis, and using our new methods that would have occurred about 26 hours after mom and dad had given consent for this protocol, and so there were two mutations that single DNA letter changes that affected a single gene, a single protein, and the diagnosis is shown there. It's a type of blood disease that happily is treatable, and so our baby was enrolled on April 10th, on April 13th, 2013, we had the answer, we called up the intensive care unit with that news, they did confirmatory tests, and then started the baby on specific treatment, precision medicines for that particular disease, and this is the promise of genomic medicine that by making a specific genetic diagnosis and understanding the cause of a child's illness, we're able to stop non-specific treatments, which are off target for that illness, and give the baby specific treatments for that condition. In this case, the story had a happy ending, and the baby's liver failure corrected itself within seven days of starting treatment. Today, this baby is 31 months old, and we estimate that we have saved 72 quality adjusted life years. This is how we measure success in our research studies. It's number one, can we make a diagnosis? Number two, can we make a diagnosis in time to change the treatment? And number three, does that have a beneficial outcome? And in this case, the baby likely would have died, and instead he now likely has a normal life expectancy, and so we've saved 72 quality life years. So the big question that our research is now addressing is whether this is true across the neonatal intensive care unit or only in a tiny subset of babies. To date, we have published just in the last year the first 35 babies that we tested, and we had an overall rate of diagnosis of 57%, which is very encouraging. And if we look across the medical literature so far, there really isn't any other literature about doing such testing in a neonatal intensive care unit, but people are doing similar studies in older children, and you can see that this agrees with the utility of these types of methods in older children, albeit it seems like the intensive care unit is an even better place to use this technology. We do need a lot more information about this since this was only 35 babies, and those babies were specifically chosen because we thought they had genetic diseases. So now we need to study and see is this true of babies in regional neonatal intensive care units, or should such testing be limited to a subset of babies? We don't yet know the answer to this. The other key question that we have, and we're continuing to answer, is in how many babies will a diagnosis change the treatment? And going into this, we were a little pessimistic. We thought that many of the babies we would be able to give mum and dad an answer, an end to the diagnostic odyssey, but maybe that it would not necessarily change the treatment of the baby in the intensive care unit. And so we also published this result this year, again it's 35 babies only, and so we need to take these results with a big pinch of salt. But so far it looks like two thirds of the babies in whom we made a diagnosis, doctors reported that it did change how they treated the baby's condition. There were four babies that were similar to the one I just described, where there was a strongly favorable impact on outcome. But as you can see, there were a lot of other changes, whether in terms of diets or medications, or the specific types of doctors who were consulted, or even in some cases, in terms of deciding that the prognosis for a baby was hopeless. And so the family started to consider about how long their baby should continue to receive care in a neonatal intensive care unit. And so our study is continuing, this is year three, and in the next year we will be continuing to enroll babies at the two intensive care units, one in Kansas City and one in San Diego, to continue to get evidence of the clinical utility of genome sequencing, and also to interview mums and dads and physicians to understand their experience and how better to provide this information. Thank you.