Big data in the age of genomics: Joel Dudley

Big data in the age of genomics: Joel Dudley



we all know that we're awash in genetic information the question is what do we do with it this morning we heard from dr. Joel Dudley of the Mount Sinai School of Medicine the Icahn Institute where he spoke to us about the resilience project and the work he's doing with big data Rolla really enjoyed your talk this morning thanks for coming out and can you tell me a little bit more about the big data that you've done yeah I think when new technologies come along whether it be various forms of genomic profiling sequencing what they allow us to do is take a new look at the world around us right it's almost like when we were first mapping out the earth we had explorers you know deciding to go into uncharted territories and sort of map out by hand what does the earth look like right but then we had satellites and now we have things like drones that can fly around the world with HD cameras and take a new look at the world around us and I think that's a bit of what genomics and molecular profiling tools let us do is to sample and seeing high-resolution the world around us and but then also the key is taking sort of a data different approach saying maybe our current knowledge is limited and these tools let us measure sort of broadly what is the picture what is the state of the world and almost how physics tries to you know discern the fundamental properties of atoms by by smashing them and collecting data and see it seen what happens to sort of understand reality we can do that a bit now in biology and that's new we haven't been able to do that that's an old idea in other fields but in biology its new which was very hypothesis drip and now we can be much more data-driven and we can find unexpected connections and the data that are just sitting right under us whether they be new uses for existing drugs or whether they be disease subtypes that we weren't aware of yeah and you pointed out some data looking at geo data set right which is a publicly available microarray data genome-wide of over a million samples then you plan an example right I visit irritable bowel disease sure yeah we point out an example of a from work I'd done previously looking at inflammatory bowel disease and epilepsy drug working for inflammatory bowel disease we found examples of antipsychotic drugs working for cancer and I think you can take that same paradigm and translate it into some things we're doing here at Mount Sinai which I talked about with our type-2 diabetes patient population by looking at our patient population in the same sort of data-driven way and looking at unexpected connections between our patients based on data we find that there might be actually multiple subpopulations of type 2 diabetes in our population and it's important because then we can find the genetic factors that would serve as markers for those subtypes so this was a lot of data then you had to compile together we showed a basically a heat map that looks something like South America yeah and another continent but of course this is data clustering and then the color indicated so the the density of the Association is correct yeah but the underlying data sets were very deep mm-hmm yeah so one thing we really believe firmly and is that in order to you know truly understand the complexity of whether it be biology whether it be a patient that really you need to combine multiple types of information you need multiple scales of data DNA is useful but that only up to a certain point RNAs you slow up to a certain point microbiome and individually they're they're powerful normally but combined they're much more powerful the idea is to paint the whole picture of what's happening in physiology and what I talked about of course is physiology and if he's multi scale but the great thing is is that we now have the molecular tools to start actually sampling not only the DNA but now the RNA the microbiome epigenetics and you can start putting these layers of information together with informatics and speaking of that then you talked a little bit about the resilience project yeah and basically looking for genetic heroes these are individuals who have a debilitating mutation they really should have be very diseased but they're actually healthy adults sure and I think this is very much in the same spirit of all the other projects whether it be drug repurposing or looking at new patient populations is that by taking a data-driven approach can we sort of take a new look a higher resolution look at everything we thought we knew and see if that's true can we question these assumptions that these these mutations are beginning a hundred percent penetrant but more importantly with the resilience project being enabled by nanotechnology that lets us you know sequence inexpensively and capture large populations what we're looking for these rare individuals that may through a quirk of nature have compensated for an extremely you know what would be in another individual devastating disease so again the technology in this case is letting us shine the white the light very wide and look at those among us to see if we happen to find a resilient individual and are they walking around with a next drug target for a disease that has no options today and then you mentioned that there is a proof of principle paper that's yet readied yes yeah so I can't do much about free published working again credit on that really goes to individuals like wrong chin and an Eric Schadt and others at Mount Sinai but so we've been able to look at you know several hundred thousand individuals under the lens of the resilience panel and I found in fact there are usually an individual among us and they're not as rare as as potentially as we thought so I'll leave yeah I can only say that well I look forward to reading that paper you know thank you again thank tell me this morning

Leave a Reply

Your email address will not be published. Required fields are marked *