Transcript

today is my last class with you a I'm sorry too you guys are a lot of fun this has actually been the most interactive 701 I've ever had and you know usually there are a couple of people who perk up and say things but you guys are great because all sorts of people are willing to contribute so I've had a wonderful time um and and it certainly seems like you guys have learned a lot what what I'd like to do for my last lecture is pick up again a little bit like I did with genomic and try to give you a sense of where things are going I always like doing this because I get to talk about things that are in none of the textbooks uh that uh well I mean that's that's just stuff that many many people working in the field don't necessarily know and that's what's so much fun about teaching introductory biology is because it only takes a semester for you guys to get up to the point of at least being able to understand what's getting done on The Cutting Edge even if you might not yet be able to go off and and practice it you might need a little more uh you know experience for that but you know you'd be surprised not that much more take take maybe project lab and you'll be able to start doing it already um it's it's really wonderful that it's possible to grasp what's going on and in many ways you guys may have an advantage in grasping what's going on because as I've already hinted biology is undergoing this remarkable transformation from from being a purely laboratory based science where each individual works on his or her own project to being a an information-based science that involves the integration of vast amounts of data across the whole world and trying to learn things from this tremendous data set and in that sense I think the new students coming into the field have a distinct advantage over those who have been in it and certainly the students who know mathematical and physical chemical and other sorts of things that aren't scared to write computer code when they need to write computer code have a really great Advantage so anyway all that by way of introduction I want to talk about two subjects today of great interest to me one is DNA variation and one is RNA variation the variation in DNA sequence uh between individuals within a population in particular our population and the other is RNA variation the variation in RNA expression between different cell types different tissues and the work I'm going to talk about today is work that that I and my colleagues have all been involved in and it's stuff I know and love so feel free to ask questions about it um I may know the answers but um what's reasonably fun about these lectures is if I don't know the answers it's probably the case that the answers aren't known so that's that's good fun because it's stuff I really do know well and and I love so anyway here's some DNA sequence um it's pretty boring uh this is a chunk of sequence from let's say the human genome how much does this differ between any two individuals if I were to sequence any two chromosomes any two copies of the chromosome from an individual in this class or two individuals on this planet how much would they differ the answer is that much that's the average amount of difference between any two people on this planet not a lot if you count it up it is on average one nucleotide difference at of 1,300 nucleotides on average somewhat less than one part in a thousand or better than 99.9% identity between any two individuals on this planet now that is a very small amount not just in you know absolute terms 99.9% identity a lot but in comparative terms with other species if I take two chimpanzees in Africa on average they will differ by about twice as much as any two random humans and if I take two orangutan in Southeast Asia they will on average differ by about eight times as much as any two humans on this planet to you guys think the orangutans all look the same they think you all look the same and they're right so why is this why are hum humans amongst Maman species uh relatively Limited in the amount of variation well it's a direct result of our population history it turns out that the amount of variation that can be sustained in a population depends on two things at equilibrium if a population has constant size n for a very long time and a certain mutation rate mu you can just write a piece of arithmetic that says well mutations are always arising due to new mutations in the population and mutations are being lost by genetic drift just by random sampling from generation to generation and those two processes the creation of new mutations and the loss of mutations just due to random sampling in each generation sets up an equilibrium and the equilibrium defines an equation there Pi = 1 1 + 4 n mu reciprocal which equation you have no need to memorize whatsoever and possibly even no need to write down the important point is the concept that if you know the number of of organisms in the population and you know the mutation rate those set up the balance of mutation and drift and you can write down how polymorphic how heterozygous random individuals should be at equilibrium that is if the population has been at size end for a very long time well the expected amount of heterozygosity for the human population sorry for a population of size 10,000 would be about one nucleotide in 1300 we have exactly the amount of heterozygosity you would expect for a population of about 10,000 individuals now but wait we're not a population of 10,000 individuals why do we have the population why do we have the heterozygosity you expect from a population of 10,000 individuals we're six billion what is it must it's a reflection of our history because remember I said that was the statement about what the population heteros should be at equilibrium we haven't been six billion people for a long you know except very recently the human population has undergone an exponential expansion it used to be a relatively small size and then very recently it underwent this huge exponential expansion if you actually write down the the equations the amount of variation in our population was determined by that constant size for a very long time and then a rapid exponential expansion that's basically taking place in a mere 3,000 Generations it's much too rapid to have any effect on the real variation in our population what do I mean by that what's the mutation rate per nucleotide in the human genome it's on the order of 2 * 10us 8 per generation in a mere 3,000 Generations a tiny mutation rate like 2 * 10us 8 is not going to be able to build up much more more variation so you might as well ignore the last 100,000 years or so they're irrelevant to how much variation we have the variation we have was set by our ancestral population size now don't get me wrong eventually eventually it will equilibrate a couple million years from now we will have a much higher variation in the human population as the function of our size but the population variation we have today is set by the fact that humans derive from a founding population of about 10,000 individuals or so and that means that the variation that you see in the human population is mostly ancestral variation it's the variation we all walked around with in Africa and in fact that makes a prediction that would say that if most of the variation in the human population is from The ancestral African founding population then if I go to any two Villages around this world in Japan or in Sweden or in Nigeria the variance that I see Will largely be identical and that prediction has been well s satisfied CU When you go and look and you collect variation in Japan or Sweden or Africa and you compare it 90% of the variants are common across the entire world most variation is common ancestral variation around the world and only a minority of the variants are new local mutations restricted to individual populations this is so contrary to what people think because there's a natural tendency to kind of xenophobia to imagine that that world populations are very different in their genetic background but in point of fact they are extremely similar so anyway there's a limited amount of variation that's why we have such little variation in the human population now that variation humans have a low rate of genetic variation most of the variants that are out there are due to common genetic variants not rare variants if I take your genome and I find a site of genetic variation at point of heterozygosity in your genome what's the probability somebody else in this class also is heterozygous for that spot turns out that the odds are about 95% that someone else in this class will also share that variant so that the variants are not mostly rare they're mostly common and it turns out that some of this common variation since it's most of this variation is likely to be important in the risk of human genetic diseases so human geneticists have gotten very excited about the following par Paradigm if there's only a limited amount of genetic variation in the human population actually if you do the arithmetic there are only about 10 million sites of common variation in the human population where common might be defined as more than about 1% in the population there only 10 million sites folks are saying well why not enumerate them all let's just know them all and then let's test each one for its risk of say conferring susceptibility of diabetes or heart disease or whatever after all 10 million is not as big a number as it used to be we now have the whole sequence of the human genome why not layer on the sequence of the human genome all common human genetic polymorphism now that's a fairly outrageous idea but could be a very useful one some of these variants are important by the way we know that there are two nucleotides that vary in the gene Apolo protein e on chromosome number 19 APO lipoprotein e it's also an APO lipoprotein like we talked about before with with familiar hyperemia but in fact it turns out that APO lipoprotein e is expressed in the brain and it turns out that amongst other tissues that it comes in three variants the spelling TT TC and CC at those two particular spots and if you happen to be homozygous for the E4 variant homozygous for the E4 variants you have about a 60 to 70% lifetime risk of Alzheimer's disease in this class 13 of you are homozygous for E4 and have a high lifetime risk of Alzheimer's and it would be fairly trivial to go across the street to anybody's lab and test that now I don't particularly recommend it and I haven't tested myself for this variant because there happens to be no particular therapy available today to delay the onset of Alzheimer's disease and therefore I don't recommend finding out about that but a number of pharmaceutical companies knowing that this is a very important Gene in the in the pathogenesis of Alzheimer's disease are working on drugs to try to delay the pathogenesis using this information and it may be the case that five or 10 years from now people begin to offer drugs that would delay the onset of Alzheimer's Disease by delaying the interaction of APO lipo protein e with a Target protein called to etc etc etc so this is an example of where a variant in the population points us to the basis of a common disease and has important therapeutic implications there are some other ones for example um 5% of you carry a particular variant in your Factor 5 Gene which is in the clotting Cascade it's called the Liden variant those 5% of you are going to account for 50% of the admissions to emergency rooms for deep Venus clots for example the much higher risk of deep Venus clots and in particular there are significant issues if you have that variant and you a woman with taking birth control pills uh some of you are at higher risk for diabetes type two adult onset diabetes there's a particular variant in the genetic Po in the population that increases your risk for type 2 diabetes by about 30% 85% of you have the high risk factor so you might as well figure you do 15% of you have a lower risk Etc and what I'm particularly interested in here it turns out that the HIV virus gets into cells with a co-receptor encoded by a gene called ccr5 well it turns out that if we go across the European population 10% of all chromosomes of European ancestry have a deletion within the C cr5 gen if 10% of all chromosomes have that deletion then 10% time 10% 1% of all individuals are homozygous for that deletion those individuals are essentially immune to infection from hi they are not susceptible it's not through immunity it's through lack of a receptor yes can you guess with as well you certainly can it's not hard it's a specific known variant you could test for it absolutely um now of course that only helps the 1% of people who are immune to who who who have that variant but what it did do was point to the pharmaceutical industry that the interaction between the virus and that variant is essential and now companies are developing drugs to block the interaction with that particular protein and that tells you you know that it's an important protein yes over the whole world I just specified your population for that one that one interestingly is not found at as higher frequency outside of Europe and no one knows why whether that might have been due to an ancient selective event or genetic drift by contrast the APO lipo protein e variant at at that prop that frequency about 3% of people being homozygous and being at risk for Alzheimer's is about the same frequency everywhere in the world so there's a little bit of population variation in frequency uh the HIV variant was found elsewhere but at considerably lower frequencies there and that's an interesting question as to what causes that that variation so you know the notion would be I've given you a couple of interesting examples but look there's only 10 million variants just write them all down make one big Excel spreadsheet with 10 million variants along the top and all the diseases along the rows and let's just fill in The Matrix and then we'll really you know this is the way people think in a post genomic era now could you do something like that you would have to enumerate all of the single nucleotide polymorphis or Snips we call them single nucleotide polymorphisms now to give you an idea of the magnitude of this problem as recently as 1998 the number of snips that were known in the human genome was a couple hundred but then project is taken off in 1998 an initial snip map of the human genome was built here at MIT that had about 4,000 of these variants then within the next year or so a inter International Consortium was organized here and elsewhere to begin to collect more of these genetic variants the goal was going to be to find 300,000 of them within a period of two years in fact that goal was blown away and within uh 3 years two million of the Snips in the human population were found and as of today if you go on the web you'll find a database with about 7.8 million of the roughly 10 million Snips in the human population already known now that isn't all 10 million and it takes a while to collect the last ones you know collecting the last ones are hard but we're already over the hump of knowing the majority of common variation in the human population not just a sequence of the genome but a database that already contains more than half of all common variation in the population so we could start building that Excel spreadsheet that turns out that it's even a little bit better than that because if we look at many chromosomes in the population here are chromosomes in the population it turns out that the common variants on each of those chromosomes tend to be correlated with each other if I know your genotype at one variant like over at this Locus I know your variant your genotype of the next Locus with reasonably high probability there's a lot of local correlation so instead of looking like a scattered picture like that it's more like this if I know that you're red red red you're probably red red red over here in other words these Vari these varations occur in blocks that we call Hao types here's real data across 111 kilobases of DNA there's a bunch of variants but it turns out that the variants come in two basic flavors 98% of all chromosomes are either this this this this this this or this this this this this this then there tends to be sites of recombination that are actually hot spots of recombination where most of the recombination of the population is concentrated and you got a couple of possibilities here so the human genome can kind of be broken up into these Hypes blocks that might be 20 30 40 sometimes 100 kilobases long in which within the block you tend to have a small number of Hypes or flavors you might think of them that uh Define most of the chromosomes in the population so in fact I don't actually need to know all the variants if they're so well correlated within a block if I knew this block structure I would be able to pick a small number of snips that would serve as a proxy for that entire block of inheritance in the population so what you might want to do is determine that entire hype block structure of how they're related to each other and pick out tag Snips and it turns out that in theory a mere 300,000 or so of them would suffice to proxy for most of the genome so you might want to declare an international project an international haa type map project to create a haa type map of the human genome and indeed such a project was declared about a year and a half ago through some instigation of scientists in a number of places including cluding here and this is aund million project uh involving count six different countries and uh it is already more than halfway done with the task and it's very likely that by the middle of next year we will have a pretty good hatip map not just show not knowing not just knowing all the variation but knowing the correlation between that variation and being able to break up the G The genome into these blocks by the next time I teach 701 I should be able to show a hype map of the whole Human Genome already that will allow you to start you're taking systematic studies of inheritance for different diseases across populations and in fact people are already doing things like that here's an example of of a study done here at MIT like this where to study inflammatory bowel disease uh there was evidence that there might be a particular region of the genome that contained it and Hypes were determined across this and blah blah blah blah blah blah blah and this red hype here turns out to confer High risk about a two and a half higher two two and a half fold higher risk of inflammatory bowel disease and it sits over some genes involved in the IM involved in immune response to certain cyto kind genes and all that and things like this have been done for type 2 diabetes schizophrenia cardiovascular disease just you know right now at the moment a dozen or two examples but I think we're set for an explosion in this kind of work in addition you can use this information to do things Beyond medical genetics you can use it for history and anthropology as well turns out rather interestingly that since the human population originated in Africa and spread out from Africa all the way around the world arriving at different parts different places and different times you can trace those migrations by virtue of rare genetic variants that arose along the way and let you like a trail of breadcrumbs see the migrations so for example there are certain rare genetic variants that we can see in a South American Indian tribe and we can actually see that they came along this route because we can see the residual of that in fact we can do things with this like take a look at Native American individuals and determine that they cluster into three distinct genetic groups that represent three distinct migrations over the land bridge and you can assign them to these different migrations you can do this on the basis of mitochondrial genotype Etc you can also for example determine when people talk about the Out of Africa migration there's now increasing evidence that there really were two one that went this way over the land and one that went this way following along the coast into southeast Asia and it looks like we're now beginning to get enough evidence of of these two separate migrations by virtue of the genetic breadcrumbs that they have left along the way so it's really a very fascinating thing of how how much you can reconstruct from looking at genetic variation both the common variation that allows us to recognize medical risk and the rare genetic variation that provides much more individual Trails of things none of this is is perfect yet there's lots to learn but I think anthropologists are finding that the existing human population has a tremendous amount of its own history embedded in the pattern of genetic variation across the world you can do other things um I probably I won't spend much time on this well I'll take a moment on this right there's some work very interesting work of a post-doctoral fellow here at MIT named Paris cetti um who has been trying to ask can we see in the genetic variation in the population signatures patterns of ancient selection or even recent selection in the human population now hang on to your seats because this will get just slightly tricky but hang on it's only a couple of slides here was her idea you see when a new mutation arises in the population it usually dies out right any new mutation typically just dies out but you know sometimes by chance it drifts up to a high frequency you know random events happen but it usually takes a long time to do that if some random mutation happens and it happens to drift up to high frequency with no selection on it then on average it takes a long time to do so if you want I could write a stochastic differential equation that would say that but you know just take your gut feeling that if something has no selection on it and it's a rare event that it'll drift up when it drifts up it's kind of a you know slow process if it was a slow process then over the course of time that it took to drift to high frequency a lot of genetic recombination would have had to have occurred many generations and the correlation between the genotype at that spot and genotype at other loai would break down and there would only be short range correlation so in other words the amount of correlation between knowing the genotype here and the genotype here maybe a Le a here and a c here that is an indication of time it's a clock almost it's like radioactive decay right that genetic recombination scrambles up the correlations and if something's old the correlations go over short distances but suppose that something happened some mutation happened that was very advantageous then it would have risen to high frequency quickly because it was under selection if it did so quickly then the long range correlations would not have had time to break down and we'd have a Smoking Gun The Smoking Gun would be that there would be a long range correlation around that Locus much longer than you would expect across the genome things even out at this distance would show correlation with that indicating that this was a recent event so we just measure across the genome and look for this Telltale sign of common variants that have very long range correlation that indicate that they're very recent so a plot of the alal frequency common variance with with long sorry if something has a common high frequency and long range correlation you wouldn't expect that by chance so something that was common in its frequency and had long range correlation would be a signature of positive selection so anyway pise had this idea and she tried it out with some interesting mutations some mutations that confer resistance to malaria one one well-known mutation causing resistance to malaria called G6PD and another one that she herself had proposed as mutation causing resistance to malaria variance in the CD4 Lian Gene and to make a long story short both the known and her newly uh predicted variant showed this Telltale property of having a high frequency and very long range correlation well that's very interesting because she was able to show that each of these mutations probably were the result of positive selection but what you could do in principle is test every variant in the human genome this way take any variant look at its frequency and compare it to the long range correlation around it and test every single variant in the human population to see which ones might be the result of long range correlation now when she proposed this was about a year year and a half ago or two years ago this was a pretty nutty idea because you would need all the variant in the human population and you would need all this correlation information but in fact as I say that information is almost upon us and I believe that this experiment this analysis to look for all all strong positive selection in the human genome will in fact be done in the course of the next 12 months so I'm hoping by next year I can actually report on a genomewide search for all the signatures of positive selection uh now this doesn't detect all positive selection it will detect sufficiently strong positive selection going back pretty much only over the last 10,000 years when you do the arithmetic that's how much power you have of course 10,000 years been a pretty interesting time for the human population right it's been you know the time of civilization and population density and infectious diseases and all that and I think we'll have a an interesting window into the select and change in diet all of that should come out of something like this so there's a lot of really cool information in DNA variation to be had all right that's one half the other half of what I would like to talk about is totally different it's not about inherited DNA variation it's about sematic differences between tissues in RNA variation so let's shift gears RNA variation let me start by giving you an example here these are cells from two different patients with acute leukemia Can you spot the difference between these yep Stu more like Bunches of grapes and all that yeah it turns out that's just the reflection of the field view you have if you move over to look like that but that's good it's just that it turns out that that isn't actually a distinction when you look at more Fields anything else yep the white blood cells look different they look broken there's more of them in this field of view but you better look at another hundred fields of view and it turns out that's not either well the reason you're having trouble spotting any difference is that highly trained Pathologists can't find any difference either um there's It generally agree there's no difference between these two if you look at enough fields of view but you can convince yourself if you look that you see things there but these actually are two very different kinds of Leukemia and these patients have to be treated very differently but Pathologists cannot determine which leukemia it is just by looking at the microscope it turns out um this is the work of this man Sydney Farber namesake of the Dana Farber Cancer Institute here in Boston who in the 1950s began noticing that patients with leukemias some of them seem different in the way they responded to a certain treatment and he said look I think there's some underlying classification of these leukemias but I can't get any reliable way to tell it in the microscope and he put many years into working this out first by noticing certain difference in enzymes in the cells and then people notice certain things in cell surface markers and some chromosomal rearrangements and nowadays there are a bunch of tests that can be done by a pathologist when a patient comes in with acute leukemia to determine whether they have AML or Al but it turns out that you can't do it by looking you have to do some kind of of a imunohistochemical test of some sort in order to do that so this is a Triumph of diagnosis uh after 40 years of work we can now correctly classify patients as AML or and they get the appropriate treatment and if they don't get the right treatment they have a much higher chance of dying and if they do get the right treatment they have a much higher chance of living so this is great there's only one problem with the story took 40 years 40 years to sort this out that's a long time could we do better surely these cells know what they are surely we could just ask them what they are well here's the idea suppose we could ask each cell please tell us every Gene that you have turned on and the and the level to which you have that Gene expressed in other words let us Summarize each cell each tumor by a description of its complete pattern of gene expression the 22,000 genes in the human genome let's write down the level of expression X1 up to X 22,000 for each of the 22,000 genes in the genome so every tumor becomes a point in 20,000 22,000 dimensional space right now clearly if we had every tumor described as a point in 22,000 dimensional space we ought to be able to sort out which tumors are similar to each other right well it turns out you can do that now these are Gene chips one of several Technologies by which on a piece of glass are put little spots each of which contains a piece of DNA a unique DNA sequence actually many copies of that DNA sequence are there each of these is a 25 Bas long DNA sequence and I can I can design this so whatever DNA sequence you want is in each spot the way that's done is with the same photo lithographic techniques that are used to make um microprocessors people have worked out of chemistry where through a mask you shine a light photo deprotect certain pixels the pixels that are photo deprotected you can chemically attach an a then reprotect the surface use a light chemically de photo deprotect certain spots wash on a c and in this fashion since you can randomly address the spots by light and then chemically add bases to whatever spots are deprotected you can simultaneously construct hundreds of thousands of spots each containing its own unique specified aligon nucleotide sequence and you can get them they're little plastic chips and then if you want all you do is you take uh a tumor you grind it up you prepare RNA you fluorescently label the RNA with some appropriate fluorescent die you squirt it into the chip you wash it back and forth you know you a rock it back and forth wash it out and stick it in a laser scanner and it'll see how much fluoresence is stuck to each spot and bingo you get a readout of the level of gene expression I guess each spot you should design it so this spot has an ol nucleotide complimentary to Gene number one and the next one an oligonucleotide matching by Crick Watson base pairing complimentary to Gene number two and Gene number three so if I knew all the genes in the genome I could make a detector spot for each gene in the genome and of course we know essentially all the genes in the genome so you can make those detector spots and you can buy them so you can now get a readout of all the I mean this is like so cool because like when I started teaching 701 which wasn't that long ago because I ain't that old still um I mean this is the way people did an analysis of gene expression is they used primitive Technologies where they would analyze one gene at a time certain things called Northern blots and things like that right and these these were you know and you you know you'd put in a lot of work and you get the expression level of a gene whereas now you can get the expression level of like all the genes simultaneously and it's you know pretty mindboggling that you can do that now how do you analyze data like that so we still use Northern plots it's true um so you know every tumor becomes a vector and we get a vector corresponding to each tumor so this is this line here is the first tumor the second tumor the third tumor the fourth tumor the The Columns here re correspond to genes there are 22,000 columns in this Matrix and I've shown a certain subset of the columns because these genes here have the interesting property that they tend to be high red in the AL tumors and they tend to be low blue in the AML tumors whereas these genes here have the opposite property they tend to be low blue in the AL tumors and high red in the AML tumors these genes do a pretty good job of telling apart these tumors so uh here's a new tumor patient came in we we analyzed that the RNA squirted on the chip can somebody classify that louder AML next next congratulations your pathologist it's very good that's right you can do that it works and in fact in the study that was done it was was published about this the computer was able to get it right 100% of the time not bad so now you say wait wait wait but you're cheating you're giving it a whole bunch of Gans once I have a whole bunch of Gans it's not so hard to classify a new tumor what Sydney Farber did was he discovered in the first place that there existed two subtypes surely that's harder than classifying when you're given a bunch of gnes and that's true so suppose uh instead I didn't tell you in advance which were amls and which were Al's and I just gave you vectors corresponding to a large number of tumors do you think you would be able to sort out that they actually fell into two clusters could you by computer tell that there's one class and the other class turns out that you can now I've made it a little easier by not listing most of the 22,000 columns here but think about it every tumor is a point in 22,000 dimensional space if some of the tumors are similar what can you say about those points in 22,000 dimensional space they're going to be clumped together they're near each other so just plot every tumor is a point in 22,000 dimensional space and your question is do the do the points tend to lie in two clumps up in 22,000 ditional space and there's simple arithmetic you you can learn using linear algebra to get some separating hyperplane and ask the tumors lie on one side or the other and it turns out that procedures like that will quickly tell you that these tumors Clump into two very clear clumps they're not randomly distributed and so if you get these tumors and you do gene expression on them and put the data into a computer the amount of time it takes the computer to discover that there are actually two types of acute leukemia is about 3 seconds marked down from 40 years it's good so you can reproduce the discovery of AML and Al in 3 seconds now you know what the Pathologists say about this they say oh give me a break it's shooting fish in a barrel we knew there was a distinction big deal that the computer can find a distinction you know we knew that there was distinction there I know you you know the computer didn't know it and all that tell us something we don't know that's a fair question so it turns out that you can ask some more questions you can say suppose I take now just the alls are they a homogeneous class or do they fall into two classes turns out that extending this work folks here were able to show that we can further split that that class there was a hint that you might be able to do so because there's some all patients who have disruptions of a gene called MLL and this tends to be a little more common in infants and tends to be associated with a poor prognosis but it was really very unclear whether this was simply one of a zillion factoids about some leukemia patients whether this was a fundamental distinction so what happened was folks took a lot of alll patients got their expression profiles and lo and behold it turned out that alll itself broke into two very different clusters this is a artist rendition of 22,000 dimensional space we can't we can't afford a 22,000 dimensional projector here so we're just using two Dimensions but um the the two forms of Al were quite distinct from each other and so actually alll itself should be split up into two classes Al plus and minus or Al one and two or MLL and Al and it turns out that these forms are quite different they have different outcomes and should be treated differently it also turns out that a particularly good distinction between Al between these two subtypes of Al is found by looking at this particular Gene called the flit 3 kyes the flit 3 kyes Gene whatever that is was of great interest because people know that they can make Inhibitors against certain kinases and and so it turned out that sorry that there is a inhibitor against flit 3 Kinesis against this fit 3 kyes Gene product if you treat cells with that inhibitor cells of this type sorry of this type die and cells of this type are not affected so in fact there's a potential drug use of fit 3 kinases in in the MLL class of these leukemias and folks are trying this in clinical trials now so not only did the analysis of the gene expression point to a two important subtypes of leukemias but the analysis of the gene expression even suggested potential targets for therapy so I can give you a bunch more examples I have a bunch more examples like that there um they are examples of taking lymphoma and showing that they can be split into two different categories examples of taking breast cancers into several categories colon cancers um basically what's going on right now is an attempt to reclassify cancers based not on what they look like in the microscope and based not on what they uh you know what what organ in the body they affect but based on molecularly what what their description is because the molecular description as Bob talked to you about with CML and with glec turns out to be a tremendously powerful way of classifying cancers because you're able to see what is the molecular defect and can make a molecular targeted therapy so these sorts of tools are quite cool and I've got to say in the last year we've begun using these expression tools not just to classify cancers but to classify drugs we beg an interesting and somewhat crazy project to take all the FDA approved drugs put them onto cell types and see what they do that is get a signature a fingerprint a gene expression description of the action of a drug and then we hope here's the Nutty idea that we could look up in the computer which drugs do which things and might be useful for which diseases because we'd put the diseases and the drugs on an equal footing all of them would be described in terms of their gene expression patterns so I'll tell you one interesting example okay this is an interesting enough example I don't even have slides for it yet turns out that these patients with all that I've been talking about some of the patients with all will respond to the drug dexamethasone some won't if you take patients who respond to dexamethasone and patients who are resistant to dexamethasone and you get their gene expression patterns you can ask are there some genes that explain the difference and you can get a certain Gene signature a list of say a dozen or so genes that do a pretty good job of classifying who's sensitive and who's resistant then you can go to this database I was telling you about of the action of many drugs and say do we see any drugs whose effect would be to produce a signature of sensitivity if we found a drug X which when we put it on cells turned on those genes that that correlate with being sensitive to dexamethasone you could hallucinate the following really happy possibility that when you added that drug together with dexamethazone you might be able to treat resistant patients because that drug could make them sensitive to dexamethasone and that you could find that drug just by looking it up in a computer database so we tried it and we hit a drug there was a certain drug that came up in the screen yes and you know that's very much in the idea too we found a drug that produced the signature of sensitivity and tested it in vitro in vitro if you take cells that are resistant and you add dexamethasone nothing happens because they're resistant if you add drug X nothing happens but if you add both drug X plus dexamethazone the cells drop dead it's now going into clinical trials in human patients turns out drug X is already a well approved you know FDA approved drugs so it can be tested in human patients right away so it's going to be tested so the gene expression pattern was able to tell us to use a drug which actually had nothing to do with cancer uses in a cancer setting because it might do something helpful now what's the point of all this um we can turn up the lights because I think I'm going to stop the slides there uh the point of all of this which is one I've made again and I will make again because you are the generation that's going to to really live this is that biology is becoming information now don't get me wrong it's not stopping being biochemistry it's going to be biochemistry that's stop stopping being molecular biology it's not stopping any of the things it was before but it is also becoming information that for the first time we're entering a world where we can collect vast amounts of information all the genetic variants in a patient all of the gene expression pattern in a cell or all of the gene expression pattern induced by a drug and that whatever question you're asking will be informed by being able to access that whole database in no way does it decrease the role of the indiv idual smart scientist working on his or her problem to the contrary the goal is to empower the individual smart scientist so that you have all of that information at your fingertips there are databases scattered around the web that have sequences from different species variations from the human population all of these drug database etc etc etc etc it's a time of tremendous ferment a little bit chaos you talk to people in the field they say we're getting deluged by data we're getting crushed by the amount of data I don't know what to do with all the data there's only one solution for a field in that condition and that is Young scientists because the young scientists who come into the field are the ones who take for granted of course we're going to have all these data we love having all these data this is just great couldn't be happier to have all these data we're not put off by it in the least that's what's going on that's what's so important about your generation and that's why I think it's really important that even though it's 701 and we're supposed to be teaching you the basics it's important that you see this stuff because this is the change that's going on and we're counting on this very much to drive a revolution in health a revolution in biomedical research and we're counting on you guys very much to drive that Revolution it has been a pleasure to teach you this term uh I hope many of you will stay in touch and some of you will go into biology and even those of you who don't will know lots about it and enjoy it thank you very much