WEBVTT 00:00:15.480 --> 00:00:17.090 PETER SZOLOVITS: So today I'm going 00:00:17.090 --> 00:00:20.010 to talk about precision medicine. 00:00:20.010 --> 00:00:22.800 And we don't really have a very precise idea 00:00:22.800 --> 00:00:25.080 of what precision medicine is. 00:00:25.080 --> 00:00:26.880 And so I'm going to start by talking 00:00:26.880 --> 00:00:29.370 about that a little bit. 00:00:29.370 --> 00:00:31.920 David talked about disease subtyping. 00:00:31.920 --> 00:00:35.280 And if you think about how do you figure out 00:00:35.280 --> 00:00:38.250 what are the subtypes of a disease, 00:00:38.250 --> 00:00:40.350 you do it by some kind of clustering 00:00:40.350 --> 00:00:42.910 on a bunch of different sorts of data. 00:00:42.910 --> 00:00:45.990 And so we have data like demographics, comorbidities, 00:00:45.990 --> 00:00:48.870 vital signs, medications, procedures, 00:00:48.870 --> 00:00:54.330 disease trajectories, whatever those mean, image similarities. 00:00:54.330 --> 00:00:57.000 And today, mostly I'm going to focus on genetics. 00:00:57.000 --> 00:01:01.020 Because this was the great hope of the Human Genome Project, 00:01:01.020 --> 00:01:04.830 that as we understood more about the genetic influences 00:01:04.830 --> 00:01:08.220 on disease, it would help us create 00:01:08.220 --> 00:01:12.390 precise ways of dealing with various diseases 00:01:12.390 --> 00:01:16.470 and figuring out the right therapies for them and so on. 00:01:16.470 --> 00:01:19.560 So I want to start by reviewing a little bit 00:01:19.560 --> 00:01:24.390 a study that was done by the National Research Council, 00:01:24.390 --> 00:01:28.470 so the National Academies, and it's called "Toward Precision 00:01:28.470 --> 00:01:30.010 Medicine." 00:01:30.010 --> 00:01:34.080 This was fairly recent, 2017. 00:01:34.080 --> 00:01:36.150 And they have some interesting observations. 00:01:36.150 --> 00:01:39.750 So they start off and they say, well, 00:01:39.750 --> 00:01:43.380 why is this relevant now, when it may not 00:01:43.380 --> 00:01:45.420 have been relevant before? 00:01:45.420 --> 00:01:48.330 And of course, the biggie is new capabilities 00:01:48.330 --> 00:01:51.870 to compile molecular data on patients on a scale that 00:01:51.870 --> 00:01:54.370 was unimaginable 20 years ago. 00:01:54.370 --> 00:01:57.960 So people estimated that getting the first human genome 00:01:57.960 --> 00:01:59.370 cost about $3 billion. 00:02:02.070 --> 00:02:07.560 Today, getting a human genome costs less than $1,000. 00:02:07.560 --> 00:02:11.730 I have some figures later in the talk showing some of the ads 00:02:11.730 --> 00:02:14.520 that people are running. 00:02:14.520 --> 00:02:17.460 Increasing success in utilizing molecular information 00:02:17.460 --> 00:02:19.830 to improve diagnosis and treatment, 00:02:19.830 --> 00:02:21.780 we'll talk about some of those. 00:02:21.780 --> 00:02:27.720 Advances in IT so that we have bigger capabilities of dealing 00:02:27.720 --> 00:02:30.690 with so-called big data-- 00:02:30.690 --> 00:02:34.170 a perfect storm among stakeholders 00:02:34.170 --> 00:02:36.270 that has made them much more receptive 00:02:36.270 --> 00:02:38.320 to this kind of information. 00:02:38.320 --> 00:02:42.630 So the fact that costs in the health-care system in the US 00:02:42.630 --> 00:02:46.920 keep rising and quality doesn't keep rising proportionately 00:02:46.920 --> 00:02:48.960 makes everybody desperate to come up 00:02:48.960 --> 00:02:52.600 with new ways of dealing with this problem. 00:02:52.600 --> 00:02:58.620 And so this looks like the next great hope for how to do it. 00:02:58.620 --> 00:03:01.780 And shifting public attitudes toward molecular data-- 00:03:01.780 --> 00:03:05.980 so how many of you have seen the movie Gattaca? 00:03:05.980 --> 00:03:06.970 A few. 00:03:06.970 --> 00:03:10.090 So that's a dystopian view of what 00:03:10.090 --> 00:03:13.960 happens when people are genotyped and can therefore 00:03:13.960 --> 00:03:16.720 be tracked by their genetics. 00:03:16.720 --> 00:03:20.620 And it is true that there are horror stories that can happen. 00:03:20.620 --> 00:03:25.600 But nevertheless, people seem to be more relaxed today 00:03:25.600 --> 00:03:29.620 about allowing that kind of data to be collected and used. 00:03:29.620 --> 00:03:34.630 Because they see the potential benefits outweighing the costs. 00:03:34.630 --> 00:03:37.990 Not everybody-- but that continues 00:03:37.990 --> 00:03:41.270 to be a serious issue. 00:03:41.270 --> 00:03:44.680 So this report goes on and says, you know, 00:03:44.680 --> 00:03:47.680 let's think about how to integrate 00:03:47.680 --> 00:03:51.190 all kinds of different data about individuals. 00:03:51.190 --> 00:03:53.170 And they start off and they say, you know, 00:03:53.170 --> 00:03:57.280 one good example of this has been Google Maps. 00:03:57.280 --> 00:04:00.350 So Google Maps has a coordinate system, 00:04:00.350 --> 00:04:03.040 which is basically longitude and latitude, 00:04:03.040 --> 00:04:05.140 for every point on the Earth. 00:04:05.140 --> 00:04:08.800 And they can use that coordinate system in order 00:04:08.800 --> 00:04:13.720 to layer on top of each other information about postal codes, 00:04:13.720 --> 00:04:17.950 built structures, census tracts, land use, transportation, 00:04:17.950 --> 00:04:19.870 everything. 00:04:19.870 --> 00:04:22.420 And they said, wow, this is really cool, if only we 00:04:22.420 --> 00:04:25.240 could do this in health care. 00:04:25.240 --> 00:04:28.950 And so their vision is to try to do that in health care 00:04:28.950 --> 00:04:31.620 by saying, well, what corresponds 00:04:31.620 --> 00:04:36.390 to latitude and longitude is individual patients. 00:04:36.390 --> 00:04:39.720 And these individual patients have various kinds 00:04:39.720 --> 00:04:42.990 of data about them, including their microbiome, 00:04:42.990 --> 00:04:47.910 their epigenome, their genome, clinical signs and symptoms, 00:04:47.910 --> 00:04:52.090 the exposome, what are they exposed to. 00:04:52.090 --> 00:04:54.630 And so there's been a real attempt 00:04:54.630 --> 00:04:59.100 to go out and create large collections of data 00:04:59.100 --> 00:05:03.160 that bring together all of this kind of information. 00:05:03.160 --> 00:05:07.560 One of those that is notable is the Department of Health-- 00:05:07.560 --> 00:05:11.280 well, NIH basically started a project about a year 00:05:11.280 --> 00:05:16.050 and a half ago called All of Us, sounds sort of menacing. 00:05:16.050 --> 00:05:19.320 But it's really a million of us. 00:05:19.320 --> 00:05:23.160 And they have asked institutions around the United States 00:05:23.160 --> 00:05:26.970 to get volunteers to volunteer to provide 00:05:26.970 --> 00:05:31.690 their genetic information, their clinical data, where they live, 00:05:31.690 --> 00:05:34.110 where they commute, things like that, 00:05:34.110 --> 00:05:37.500 so that they can get environmental data about them. 00:05:37.500 --> 00:05:41.790 And then it's meant to be an ongoing collection of data 00:05:41.790 --> 00:05:44.850 about a million people who are supposed 00:05:44.850 --> 00:05:48.360 to be a representative sample of the United States. 00:05:48.360 --> 00:05:50.550 So you'll see in some of the projects I talk 00:05:50.550 --> 00:05:53.310 about later today that many of the studies 00:05:53.310 --> 00:05:57.130 have been done in populations of European ancestry. 00:05:57.130 --> 00:06:00.930 And so they may not apply to people of other ethnicities. 00:06:00.930 --> 00:06:05.370 This is attempting to sample accurately so 00:06:05.370 --> 00:06:09.120 that the fraction of African Americans and Asians 00:06:09.120 --> 00:06:12.570 and Hispanics and so on corresponds 00:06:12.570 --> 00:06:16.180 to the sample in the United States population. 00:06:16.180 --> 00:06:17.950 There's a long history. 00:06:17.950 --> 00:06:21.120 How many of you have heard of the Framingham Heart Study? 00:06:21.120 --> 00:06:22.350 So a lot of people. 00:06:22.350 --> 00:06:25.710 So Framingham, in the 1940s, agreed 00:06:25.710 --> 00:06:29.370 to become the subject of a long-term experiment. 00:06:29.370 --> 00:06:35.340 I think it's now run by Boston University, where every year 00:06:35.340 --> 00:06:40.520 or two they go out and they survey-- 00:06:40.520 --> 00:06:42.510 I can't remember the number of people. 00:06:42.510 --> 00:06:47.370 It started as something like 50,000 people-- 00:06:47.370 --> 00:06:49.920 about their habits and whether they smoke, 00:06:49.920 --> 00:06:52.020 and what their weight and height is, 00:06:52.020 --> 00:06:54.870 and any clinical problems they've had, 00:06:54.870 --> 00:06:57.300 surgical procedures, et cetera. 00:06:57.300 --> 00:06:59.580 And they've been collecting that database now 00:06:59.580 --> 00:07:05.160 over several generations of people that descend from those. 00:07:05.160 --> 00:07:08.260 And they've started collecting genetic data as well. 00:07:08.260 --> 00:07:12.810 So All of Us is really doing this on a very large scale. 00:07:12.810 --> 00:07:16.560 Now, the vision of these of this study 00:07:16.560 --> 00:07:20.370 was to say that we're going to build this information 00:07:20.370 --> 00:07:24.270 commons, which collects all this kind of information, 00:07:24.270 --> 00:07:29.100 and then we're going to develop knowledge from that information 00:07:29.100 --> 00:07:30.780 or from that data. 00:07:30.780 --> 00:07:33.900 And that knowledge will become the substrate on which 00:07:33.900 --> 00:07:36.340 biomedical research can rest. 00:07:36.340 --> 00:07:39.510 So if we find significant associations, 00:07:39.510 --> 00:07:41.310 then that suggests that one should 00:07:41.310 --> 00:07:44.190 do studies, which will not necessarily 00:07:44.190 --> 00:07:46.950 be answered by the data that they've collected. 00:07:46.950 --> 00:07:49.890 You may have to grow knock-out mice or something in order 00:07:49.890 --> 00:07:52.560 to test whether an idea really works. 00:07:52.560 --> 00:07:54.660 But this is a way of integrating all 00:07:54.660 --> 00:07:56.920 of that type of information. 00:07:56.920 --> 00:07:59.530 And of course, it can affect diagnosis, treatment, 00:07:59.530 --> 00:08:02.490 and health outcomes, which are the holy grail for what 00:08:02.490 --> 00:08:05.880 you'd like to do in medicine. 00:08:05.880 --> 00:08:09.060 Now, here's an interesting problem. 00:08:09.060 --> 00:08:14.310 So the focus, notice, was on taxonomies. 00:08:14.310 --> 00:08:21.540 So Sam Johnson was a very famous 17th century British writer. 00:08:21.540 --> 00:08:26.310 And he built encyclopedias and dictionaries, 00:08:26.310 --> 00:08:30.690 and was a poet and a reviewer and a commentator, 00:08:30.690 --> 00:08:34.200 and did all kinds of fancy things. 00:08:34.200 --> 00:08:36.900 And one of his quotes is, "My diseases 00:08:36.900 --> 00:08:40.380 are an asthma and a dropsy and, what is less curable, 00:08:40.380 --> 00:08:42.820 75," years old. 00:08:42.820 --> 00:08:46.380 So he was funny, too. 00:08:46.380 --> 00:08:52.020 Now, if you look up dropsy in a dictionary-- 00:08:52.020 --> 00:08:55.370 how many of you have heard of dropsy? 00:08:55.370 --> 00:08:56.570 A couple. 00:08:56.570 --> 00:08:59.094 So how did you hear of it? 00:08:59.094 --> 00:09:01.310 AUDIENCE: From Jane Austen novels. 00:09:01.310 --> 00:09:01.810 [LAUGHS] 00:09:01.810 --> 00:09:02.768 PETER SZOLOVITS: Sorry? 00:09:02.768 --> 00:09:03.470 From a novel? 00:09:03.470 --> 00:09:04.490 AUDIENCE: Novels. 00:09:04.490 --> 00:09:05.407 PETER SZOLOVITS: Yeah. 00:09:05.407 --> 00:09:08.043 AUDIENCE: I've heard of dropsy [INAUDIBLE].. 00:09:08.043 --> 00:09:08.960 PETER SZOLOVITS: Yeah. 00:09:08.960 --> 00:09:12.380 I mean, I learned about it by watching Masterpiece Theatre 00:09:12.380 --> 00:09:18.200 with 19th century people, where the grandmother would take 00:09:18.200 --> 00:09:21.260 to her bed with the dropsy. 00:09:21.260 --> 00:09:23.870 And it didn't turn out well, typically. 00:09:23.870 --> 00:09:27.050 But it took a long time for those people to die. 00:09:27.050 --> 00:09:31.010 So dropsy is water sickness, swelling, edema, et cetera. 00:09:31.010 --> 00:09:32.910 It's actually not a disease. 00:09:32.910 --> 00:09:35.720 It's a symptom of a whole bunch of diseases. 00:09:35.720 --> 00:09:39.540 So it could be pulmonary disease, heart failure, 00:09:39.540 --> 00:09:42.200 kidney disease, et cetera. 00:09:42.200 --> 00:09:43.220 And it's interesting. 00:09:43.220 --> 00:09:44.360 I look back on this. 00:09:44.360 --> 00:09:47.840 I couldn't find it for putting together this lecture. 00:09:47.840 --> 00:09:52.190 But at one point, I did discover that the last time dropsy was 00:09:52.190 --> 00:09:56.060 listed as the cause of death of a patient in the United States 00:09:56.060 --> 00:09:58.520 was in 1949. 00:09:58.520 --> 00:10:04.100 So since then, it's disappeared as a disease from the taxonomy. 00:10:04.100 --> 00:10:07.070 And if you talk to pulmonary people, 00:10:07.070 --> 00:10:09.950 they suspect that asthma, which is still 00:10:09.950 --> 00:10:13.770 a disease in our current lexicon, 00:10:13.770 --> 00:10:15.890 may be very much like dropsy. 00:10:15.890 --> 00:10:17.250 It's not a disease. 00:10:17.250 --> 00:10:21.050 It's a symptom of a whole bunch of underlying causes. 00:10:21.050 --> 00:10:22.940 And the idea is that we need to get 00:10:22.940 --> 00:10:27.110 good enough and precise enough at being able to figure out 00:10:27.110 --> 00:10:29.670 what these are. 00:10:29.670 --> 00:10:35.300 So I talked to my friend Zack Kohane at Harvard 00:10:35.300 --> 00:10:38.780 a few weeks ago when I started preparing this lecture. 00:10:38.780 --> 00:10:41.480 And he has the following idea. 00:10:41.480 --> 00:10:46.080 And the example I'm going to show you is from him. 00:10:46.080 --> 00:10:48.110 So he says, well, look, we should 00:10:48.110 --> 00:10:50.930 have this precision medicine modality 00:10:50.930 --> 00:10:54.920 space, which is this high-dimensional space that 00:10:54.920 --> 00:11:01.160 contains all of that information that is in the NRC report. 00:11:01.160 --> 00:11:07.880 And then what we do is, in this high-dimensional space, 00:11:07.880 --> 00:11:11.630 if we're lucky, we're going to find clusters of data. 00:11:11.630 --> 00:11:14.030 So this always happens. 00:11:14.030 --> 00:11:17.960 If you ever take a very high-dimensional data set 00:11:17.960 --> 00:11:21.800 and put it into its very high-dimensional representation 00:11:21.800 --> 00:11:24.950 space, it's almost never the case 00:11:24.950 --> 00:11:30.320 that the data is scattered uniformly through the space. 00:11:30.320 --> 00:11:33.350 If that were true, it wouldn't help us very much. 00:11:33.350 --> 00:11:35.450 But generally, it's not true. 00:11:35.450 --> 00:11:37.160 And what you find is that the data 00:11:37.160 --> 00:11:40.790 tends to be on lower-dimensional manifolds. 00:11:40.790 --> 00:11:43.650 So it's in subsets of the space. 00:11:43.650 --> 00:11:45.800 And so a lot of the trick in trying 00:11:45.800 --> 00:11:49.040 to analyze this kind of data is figuring out 00:11:49.040 --> 00:11:53.000 what those lower-dimensional manifolds look like. 00:11:53.000 --> 00:11:58.220 And often you will find among a very large data set a cluster 00:11:58.220 --> 00:12:00.920 of patients like this. 00:12:00.920 --> 00:12:06.020 And then Zack's approach is to say, well, if you're patient-- 00:12:06.020 --> 00:12:08.810 it's hard to represent three dimensions in two. 00:12:08.810 --> 00:12:11.780 But if you're patient that falls somewhere in the middle of such 00:12:11.780 --> 00:12:14.420 a cluster, then that probably means 00:12:14.420 --> 00:12:17.630 that they're kind of normal for that cluster, 00:12:17.630 --> 00:12:21.500 whereas if they fall somewhere at the edge of such a cluster, 00:12:21.500 --> 00:12:23.660 that probably means that there's something odd 00:12:23.660 --> 00:12:26.900 going on that is worth investigating, because they're 00:12:26.900 --> 00:12:28.820 unusual. 00:12:28.820 --> 00:12:34.370 So then he gave me an example of a patient of his. 00:12:34.370 --> 00:12:37.570 And let me give you a minute to read this. 00:12:50.902 --> 00:12:51.402 Yeah? 00:12:51.402 --> 00:12:53.460 AUDIENCE: What's an armamentarium? 00:12:56.893 --> 00:12:58.935 PETER SZOLOVITS: Where does it say armamentarium? 00:12:58.935 --> 00:12:59.730 AUDIENCE: [INAUDIBLE] 00:12:59.730 --> 00:13:00.813 PETER SZOLOVITS: Oh, yeah. 00:13:00.813 --> 00:13:03.810 So an armamentarium, historically, 00:13:03.810 --> 00:13:07.230 is the set of arms that are available to an army. 00:13:07.230 --> 00:13:10.560 So this is the set of treatments that are available to a doctor. 00:13:13.050 --> 00:13:15.008 AUDIENCE: Is that the only word you don't know? 00:13:15.008 --> 00:13:17.378 [LAUGHTER] 00:13:17.378 --> 00:13:18.330 It's the only word-- 00:13:18.330 --> 00:13:19.960 AUDIENCE: If I start asking-- 00:13:19.960 --> 00:13:21.252 AUDIENCE: Based on [INAUDIBLE]. 00:13:21.252 --> 00:13:22.218 AUDIENCE: Oh, OK. 00:13:22.218 --> 00:13:24.150 AUDIENCE: In the world. 00:13:24.150 --> 00:13:26.570 Some of it, I thought I could understand. 00:13:26.570 --> 00:13:28.028 PETER SZOLOVITS: Well, you probably 00:13:28.028 --> 00:13:29.750 know what antibiotics are. 00:13:29.750 --> 00:13:33.440 And immunosuppressants, you've probably heard of. 00:13:33.440 --> 00:13:37.820 Anyway, it's a bunch of different therapies. 00:13:37.820 --> 00:13:41.270 So this is what's called a sick puppy. 00:13:41.270 --> 00:13:44.900 It's a kid who is not doing well. 00:13:44.900 --> 00:13:49.220 They started life, at age three, with ulcerative colitis, 00:13:49.220 --> 00:13:52.010 which was well-controlled by the kinds of medications 00:13:52.010 --> 00:13:56.090 that they normally give people with that disease. 00:13:56.090 --> 00:13:59.570 And then all of a sudden, 10 years later, 00:13:59.570 --> 00:14:03.710 he breaks out with this horrible abdominal pain and diarrhea 00:14:03.710 --> 00:14:07.490 and blood in his stool. 00:14:07.490 --> 00:14:12.717 And they try a bunch of stuff that they think ought to work, 00:14:12.717 --> 00:14:13.550 and it doesn't work. 00:14:16.220 --> 00:14:23.030 So the kid was facing some fairly drastic options, 00:14:23.030 --> 00:14:28.410 like cutting out the part of his colon that was inflamed. 00:14:28.410 --> 00:14:31.950 So your colon is an important part of your digestive tract. 00:14:31.950 --> 00:14:36.770 And so losing it is not fun and would have bad consequences 00:14:36.770 --> 00:14:40.520 for the rest of his life. 00:14:40.520 --> 00:14:46.850 But what they did is they said, well, 00:14:46.850 --> 00:14:51.790 why is he not responding to any of these therapies? 00:14:51.790 --> 00:14:57.330 And the difficulty, you can imagine, 00:14:57.330 --> 00:15:01.050 in that cloud-of-points picture, is, 00:15:01.050 --> 00:15:04.860 how do you figure out whether the person is an outlier 00:15:04.860 --> 00:15:07.820 or is in the middle of one of these clusters, 00:15:07.820 --> 00:15:09.870 when it depends on a lot of things? 00:15:09.870 --> 00:15:13.560 In this kid's case, what it depended on most significantly 00:15:13.560 --> 00:15:16.920 was the last six months of his experience, 00:15:16.920 --> 00:15:21.980 where, before, he was doing OK with the standard treatment. 00:15:21.980 --> 00:15:24.830 So that cloud might have represented people 00:15:24.830 --> 00:15:27.650 with ulcerative colitis who were well-controlled 00:15:27.650 --> 00:15:29.330 by the standard treatment. 00:15:29.330 --> 00:15:33.360 And now, all of a sudden, he becomes an outlier. 00:15:33.360 --> 00:15:38.850 So what happened in this case is they said, 00:15:38.850 --> 00:15:42.170 well, maybe there are different groups 00:15:42.170 --> 00:15:44.420 of ulcerative colitis patients. 00:15:44.420 --> 00:15:48.080 So maybe there are ones who have a lifelong remission 00:15:48.080 --> 00:15:51.260 after treatment with a commonly used monoclonal antibody. 00:15:51.260 --> 00:15:56.090 So that's the center of the cluster. 00:15:56.090 --> 00:15:59.540 Maybe there are people who have multi-year remission 00:15:59.540 --> 00:16:03.230 but become refractory to these drugs. 00:16:03.230 --> 00:16:09.200 And after other treatments, they have to undergo a colectomy. 00:16:09.200 --> 00:16:12.560 So that's the removal of the colon. 00:16:12.560 --> 00:16:15.590 And then there are people who have, initially, a remission, 00:16:15.590 --> 00:16:18.230 but then those standard therapy works. 00:16:18.230 --> 00:16:22.550 So that's what this kid is in, this cluster. 00:16:22.550 --> 00:16:29.470 So how do you treat this as a machine learning problem 00:16:29.470 --> 00:16:31.750 from the point of view of having lots of data 00:16:31.750 --> 00:16:34.060 about lots of different patients? 00:16:34.060 --> 00:16:38.560 And the challenges, of course, include things like, 00:16:38.560 --> 00:16:42.250 what's your distance function in doing the kind of clustering 00:16:42.250 --> 00:16:44.650 that people typically do? 00:16:44.650 --> 00:16:47.920 How do you define what an outlier is? 00:16:47.920 --> 00:16:51.280 Because there's always a continuum where it just 00:16:51.280 --> 00:16:54.190 gets more and more diffuse. 00:16:54.190 --> 00:16:57.610 What's the best representation for time-varying data, 00:16:57.610 --> 00:17:00.160 which is critical in this case? 00:17:00.160 --> 00:17:04.069 What's the optimal weighting or normalization of dimensions? 00:17:04.069 --> 00:17:07.630 So does every dimension in this very high-dimensional space 00:17:07.630 --> 00:17:09.040 count the same? 00:17:09.040 --> 00:17:11.380 Or are differences along certain dimensions 00:17:11.380 --> 00:17:13.990 more important than those among others? 00:17:13.990 --> 00:17:17.900 And does that, in fact, vary from problem to problem? 00:17:17.900 --> 00:17:21.160 The answer is probably yes. 00:17:21.160 --> 00:17:25.300 So how do we find the neighborhood for the patient? 00:17:25.300 --> 00:17:29.110 Well, I'm going to give you some clues 00:17:29.110 --> 00:17:32.590 by starting with a shallow dive into genetics. 00:17:32.590 --> 00:17:37.390 So if you've taken a molecular cell biology class, 00:17:37.390 --> 00:17:39.080 this should not be news to you. 00:17:39.080 --> 00:17:41.320 And I'm going to run through it pretty quickly. 00:17:41.320 --> 00:17:43.480 If you haven't, then I hope at least 00:17:43.480 --> 00:17:46.120 you'll pick up some of the vocabulary. 00:17:46.120 --> 00:17:49.240 So a wise biologist said, "Biology 00:17:49.240 --> 00:17:51.670 is the science of exceptions." 00:17:51.670 --> 00:17:53.990 There are almost no rules. 00:17:53.990 --> 00:17:57.820 About 25 years ago, the biology department 00:17:57.820 --> 00:18:02.590 here taught a special class for engineering faculty 00:18:02.590 --> 00:18:05.320 to try to explain to us what they 00:18:05.320 --> 00:18:08.150 were teaching in their introductory biology, 00:18:08.150 --> 00:18:10.240 molecular biology classes. 00:18:10.240 --> 00:18:12.400 And I remember, I was sitting next 00:18:12.400 --> 00:18:14.800 to Jerry Sussman, one of my colleagues. 00:18:14.800 --> 00:18:19.150 And after we heard some lecture about the 47 ways 00:18:19.150 --> 00:18:23.460 that some theory doesn't apply in many, many cases, 00:18:23.460 --> 00:18:25.780 Jerry turns to me and he says, you know, 00:18:25.780 --> 00:18:28.540 the problem with this field is there are just too many damned 00:18:28.540 --> 00:18:29.950 exceptions. 00:18:29.950 --> 00:18:31.990 There are no theories. 00:18:31.990 --> 00:18:34.540 It's all exceptions. 00:18:34.540 --> 00:18:39.310 And so even biologists recognize this. 00:18:39.310 --> 00:18:43.570 Now, people have observed, ever since human beings walked 00:18:43.570 --> 00:18:45.610 the earth, that children tend to be 00:18:45.610 --> 00:18:49.030 similar to their parents in many ways. 00:18:49.030 --> 00:18:53.470 And until Gregor Mendel, this was a great mystery. 00:18:53.470 --> 00:18:57.077 Why is it that you are like your parents? 00:18:57.077 --> 00:18:58.660 I mean, you must have gotten something 00:18:58.660 --> 00:19:01.750 from them that sort of carries through and makes 00:19:01.750 --> 00:19:03.790 you similar to them. 00:19:03.790 --> 00:19:06.880 So Mendel had this notion of having 00:19:06.880 --> 00:19:10.810 discrete factors of inheritance, which he called genes. 00:19:10.810 --> 00:19:13.060 He had no idea what these were. 00:19:13.060 --> 00:19:17.080 But conceptually, he knew that they must exist. 00:19:17.080 --> 00:19:20.860 And then he did a bunch of experiments on pea plants, 00:19:20.860 --> 00:19:23.920 showing that peas that are wrinkled 00:19:23.920 --> 00:19:28.330 tend to have offspring peas that are also wrinkled. 00:19:28.330 --> 00:19:33.010 And he worked out the genetics of what we now 00:19:33.010 --> 00:19:36.430 call Mendelian inheritance, namely 00:19:36.430 --> 00:19:41.710 dominant versus recessive inheritance patterns. 00:19:41.710 --> 00:19:47.980 Then Johann Miescher came along some years later, 00:19:47.980 --> 00:19:52.270 and he discovered a weird thing in cells 00:19:52.270 --> 00:19:56.545 called nuclein, which is now known as DNA. 00:20:00.220 --> 00:20:07.940 But it wasn't until 1952 that Hershey and Chase said, hey, 00:20:07.940 --> 00:20:11.950 it's DNA that is carrying this genetic information 00:20:11.950 --> 00:20:15.400 from generation to generation. 00:20:15.400 --> 00:20:18.100 And then, of course, Watson, Crick, and Franklin, 00:20:18.100 --> 00:20:22.120 the following year, deciphered the structure of DNA, 00:20:22.120 --> 00:20:25.180 that it's this double helix, and then figured out 00:20:25.180 --> 00:20:28.480 what the mechanism must be that allows DNA 00:20:28.480 --> 00:20:30.920 to transmit this information. 00:20:30.920 --> 00:20:32.840 So you have a double helix. 00:20:32.840 --> 00:20:44.570 You match the four letters A, C, T, G opposite each other, 00:20:44.570 --> 00:20:47.380 and you can replicate this DNA by splitting it 00:20:47.380 --> 00:20:50.050 apart and growing another strand that 00:20:50.050 --> 00:20:52.430 is the complement of the first one. 00:20:52.430 --> 00:20:53.470 Now you have two. 00:20:53.470 --> 00:20:57.980 And you can have children, pass on this information to them. 00:20:57.980 --> 00:21:00.110 So that was a big deal. 00:21:00.110 --> 00:21:04.330 So a gene is defined by the National Center 00:21:04.330 --> 00:21:07.930 for Biotechnology Information as a fundamental physical 00:21:07.930 --> 00:21:11.980 and functional unit of heredity that's a DNA sequence located 00:21:11.980 --> 00:21:14.830 on a specific site on a chromosome which 00:21:14.830 --> 00:21:17.590 encodes a specific functional product, 00:21:17.590 --> 00:21:20.350 namely RNA or a protein. 00:21:20.350 --> 00:21:22.710 I'll come back to that in a minute. 00:21:22.710 --> 00:21:24.940 The remaining mystery is it's still very 00:21:24.940 --> 00:21:29.980 hard to figure out what parts of the DNA code genes. 00:21:29.980 --> 00:21:32.090 So you would think we might have solved this, 00:21:32.090 --> 00:21:34.210 but we haven't quite. 00:21:34.210 --> 00:21:37.840 And what does the rest, which is the vast majority of the DNA, 00:21:37.840 --> 00:21:41.920 do if it's not encoding genes? 00:21:41.920 --> 00:21:45.250 And then, how does the folding and the geometry, 00:21:45.250 --> 00:21:49.383 the topology of these structures, 00:21:49.383 --> 00:21:50.425 influence their function? 00:21:53.270 --> 00:21:57.250 So I went back and I read some of Francis Crick's work 00:21:57.250 --> 00:21:59.660 from the 1950s. 00:21:59.660 --> 00:22:01.360 And it's very interesting. 00:22:01.360 --> 00:22:06.130 This hypothesis was considered controversial and tentative 00:22:06.130 --> 00:22:07.420 at the time. 00:22:07.420 --> 00:22:11.290 So he said, "The specificity of a piece of nucleic acid 00:22:11.290 --> 00:22:14.950 is expressed solely by the sequence of its bases, 00:22:14.950 --> 00:22:16.960 and this sequence is a simple code 00:22:16.960 --> 00:22:21.010 for the amino acid sequence of a particular protein." 00:22:21.010 --> 00:22:24.190 And there were people arguing that he was just flat wrong, 00:22:24.190 --> 00:22:25.450 that this was not true. 00:22:25.450 --> 00:22:28.900 Of course, it turned out he was right. 00:22:28.900 --> 00:22:31.480 And then the central dogma is the transfer 00:22:31.480 --> 00:22:35.110 of information from nucleic acid to nucleic acid 00:22:35.110 --> 00:22:38.810 or from nucleic acid to protein may be possible. 00:22:38.810 --> 00:22:42.790 But transfer from protein to protein or from protein 00:22:42.790 --> 00:22:45.820 to nucleic acid is impossible. 00:22:45.820 --> 00:22:48.160 And that's not quite true. 00:22:48.160 --> 00:22:51.380 But it's a good first approximation. 00:22:51.380 --> 00:22:57.740 So this is where things stood back about 60 years ago. 00:22:57.740 --> 00:23:00.800 And then a few Nobel prizes later, 00:23:00.800 --> 00:23:03.890 we began to understand some of the mechanism of how 00:23:03.890 --> 00:23:05.180 this works. 00:23:05.180 --> 00:23:07.490 And of course, how it works is that you 00:23:07.490 --> 00:23:13.250 have DNA, which is these four bases, double stranded. 00:23:13.250 --> 00:23:17.400 RNA gets produced in the process of transcription. 00:23:17.400 --> 00:23:20.720 So this thing unfolds. 00:23:20.720 --> 00:23:27.710 An RNA strand is built along the DNA and separates from the DNA, 00:23:27.710 --> 00:23:30.380 creating a single-stranded RNA. 00:23:30.380 --> 00:23:33.320 And then it goes and hooks up with a ribosome. 00:23:33.320 --> 00:23:38.990 And the ribosome takes that RNA and takes the codes 00:23:38.990 --> 00:23:42.020 in triplets, and each triplet stands 00:23:42.020 --> 00:23:45.710 for a particular amino acid, which it then assembles 00:23:45.710 --> 00:23:48.020 in sequence and creates proteins, which 00:23:48.020 --> 00:23:52.300 are sequences of amino acids. 00:23:52.300 --> 00:23:57.210 Now, it's very complicated. 00:23:57.210 --> 00:23:59.340 Because there's three-dimensionality 00:23:59.340 --> 00:24:00.720 and there's time involved. 00:24:00.720 --> 00:24:05.310 And the rate constants-- this is chemistry, after all. 00:24:05.310 --> 00:24:10.245 So again, a few more Nobel prizes later, 00:24:10.245 --> 00:24:15.330 we have that transcription, that process of turning DNA 00:24:15.330 --> 00:24:19.140 into RNA, is regulated by promoter, repressor, 00:24:19.140 --> 00:24:22.860 and enhancer regions on the genome. 00:24:22.860 --> 00:24:27.960 And the proteins mediate this process by binding to the DNA 00:24:27.960 --> 00:24:31.560 and causing the beginning of transcription, 00:24:31.560 --> 00:24:35.340 or causing it to run faster or causing it to run slower, 00:24:35.340 --> 00:24:38.730 or they interfere with it, et cetera. 00:24:38.730 --> 00:24:42.000 There are also these enhancers, some of which 00:24:42.000 --> 00:24:47.280 are very far away from the coding region, that 00:24:47.280 --> 00:24:51.210 make huge differences in how much of the RNA, 00:24:51.210 --> 00:24:54.510 and therefore how much of the protein, is made. 00:24:54.510 --> 00:24:57.270 And the current understanding of that 00:24:57.270 --> 00:25:00.360 is that, if here is the gene, it may 00:25:00.360 --> 00:25:03.750 be that the strand of DNA loops around. 00:25:03.750 --> 00:25:08.700 And the enhancer, even though it's distant in genetic units, 00:25:08.700 --> 00:25:12.130 is actually in close physical proximity, 00:25:12.130 --> 00:25:16.860 and therefore can encourage more of this transcription 00:25:16.860 --> 00:25:20.160 to take place. 00:25:20.160 --> 00:25:23.550 By the way, if you're interested in this stuff, of course 00:25:23.550 --> 00:25:28.230 MIT teaches a lot of courses in how to do this. 00:25:28.230 --> 00:25:31.860 Dave Gifford and Manolis Kellis both teach 00:25:31.860 --> 00:25:36.000 computational courses in how to apply computational methods 00:25:36.000 --> 00:25:39.820 to try to decipher this kind of activity. 00:25:39.820 --> 00:25:43.470 So repressors prevent activator from binding or alters 00:25:43.470 --> 00:25:46.770 the activator in order to change the rate constants. 00:25:46.770 --> 00:25:50.340 And so this is another mechanism. 00:25:50.340 --> 00:25:53.930 Now, one of the problems is that if you 00:25:53.930 --> 00:26:02.510 look at the total amount of DNA in your genes, in your cells, 00:26:02.510 --> 00:26:07.160 only about 1 and 1/2% are exons, which 00:26:07.160 --> 00:26:14.820 are the parts that code for mRNA, and eventually protein. 00:26:14.820 --> 00:26:19.600 So the question is what does the other 98 and 1/2% do? 00:26:19.600 --> 00:26:23.650 There was this unfortunate tendency in the biology 00:26:23.650 --> 00:26:27.100 community to call that junk DNA, which 00:26:27.100 --> 00:26:29.200 of course is a terrible notion. 00:26:29.200 --> 00:26:31.990 Because evolution would certainly have gotten 00:26:31.990 --> 00:26:34.540 rid of it if it was truly junk. 00:26:34.540 --> 00:26:39.710 Because our cells spend a lot of energy building this stuff. 00:26:39.710 --> 00:26:42.700 And every time a cell divides, it 00:26:42.700 --> 00:26:45.640 rebuilds all that so-called junk DNA. 00:26:45.640 --> 00:26:48.700 So it can't possibly be junk. 00:26:48.700 --> 00:26:51.280 But the question is, what does it do? 00:26:51.280 --> 00:26:54.890 And we don't really know for a lot of it. 00:26:54.890 --> 00:26:57.370 So there are introns-- 00:26:57.370 --> 00:26:58.580 I'll show you a picture. 00:26:58.580 --> 00:27:02.830 There are segments of the coding region that don't wind up 00:27:02.830 --> 00:27:04.180 as part of the RNA. 00:27:04.180 --> 00:27:05.890 They're spliced out. 00:27:05.890 --> 00:27:08.650 And we don't quite know why. 00:27:08.650 --> 00:27:11.350 There are these regulatory sequences, 00:27:11.350 --> 00:27:15.070 which is only about 5%, that are those promoters 00:27:15.070 --> 00:27:22.880 and repressors and enhancers that I talked about. 00:27:22.880 --> 00:27:26.230 And then there's a whole bunch of repetitive DNA that 00:27:26.230 --> 00:27:30.980 includes transposable elements, related sequences. 00:27:30.980 --> 00:27:34.150 And mostly, we don't understand what it all does. 00:27:36.880 --> 00:27:39.700 Hypotheses are things like, well, maybe 00:27:39.700 --> 00:27:43.870 it's a storehouse of potentially useful DNA 00:27:43.870 --> 00:27:47.750 so that if environmental conditions change a lot, 00:27:47.750 --> 00:27:49.720 then the cell doesn't have to reinvent 00:27:49.720 --> 00:27:51.490 the stuff from scratch. 00:27:51.490 --> 00:27:55.740 It saved it from previous times in evolution 00:27:55.740 --> 00:27:57.460 when that may have been useful. 00:27:57.460 --> 00:28:01.640 But that's pretty much pure speculation at this point. 00:28:01.640 --> 00:28:04.300 So just recently, the Killian Lecture 00:28:04.300 --> 00:28:09.520 was given by Gerald Fink, who's a geneticist here. 00:28:09.520 --> 00:28:14.590 And his claim is that a gene is not any segment of DNA 00:28:14.590 --> 00:28:17.260 that produces RNA or protein. 00:28:17.260 --> 00:28:20.770 But it's any segment of DNA that is transcribed 00:28:20.770 --> 00:28:23.740 into RNA that has some function, whatever 00:28:23.740 --> 00:28:29.060 it is, not necessarily building proteins, but just anything. 00:28:29.060 --> 00:28:33.200 And I think that view is becoming accepted. 00:28:33.200 --> 00:28:39.130 So I promised you a little bit of more complexity. 00:28:39.130 --> 00:28:42.520 So when you look at your DNA in eukaryotes, 00:28:42.520 --> 00:28:45.820 like us, here's the promoter. 00:28:45.820 --> 00:28:51.070 And then here is the sequence of the genome. 00:28:51.070 --> 00:28:55.420 When this gets transcribed, it gets transcribed into something 00:28:55.420 --> 00:28:59.320 called pre-mRNA, messenger RNA. 00:28:59.320 --> 00:29:03.120 And then there's this process of alternative splicing 00:29:03.120 --> 00:29:09.100 that splices out the introns and leaves only the exons. 00:29:09.100 --> 00:29:11.620 But sometimes it doesn't leave all the exons. 00:29:11.620 --> 00:29:13.640 It only leaves some of them. 00:29:13.640 --> 00:29:17.470 And so the same gene can, under various circumstances, 00:29:17.470 --> 00:29:19.840 produce different mRNA, which then 00:29:19.840 --> 00:29:22.310 produces different proteins. 00:29:22.310 --> 00:29:25.540 And again, there's a lot of mysteries about exactly 00:29:25.540 --> 00:29:27.940 how all this works. 00:29:27.940 --> 00:29:31.210 Nevertheless, that's the basic mechanism. 00:29:31.210 --> 00:29:36.325 And then here, I've just listed a few 00:29:36.325 --> 00:29:38.860 of the complexity problems. 00:29:38.860 --> 00:29:43.630 So there are things like, RNA can turn into DNA. 00:29:43.630 --> 00:29:46.810 This is a trick that viruses use a lot. 00:29:46.810 --> 00:29:50.260 They incorporate themselves into your cell, 00:29:50.260 --> 00:29:55.390 create a DNA complement to the RNA, 00:29:55.390 --> 00:29:58.670 and then use that to generate more viruses. 00:29:58.670 --> 00:30:03.340 So this is very typical of a viral infection. 00:30:03.340 --> 00:30:06.700 Prions, we also don't understand very well. 00:30:06.700 --> 00:30:11.740 This is like mad cow disease, where these proteins are able 00:30:11.740 --> 00:30:16.000 to cause changes in other proteins without going through 00:30:16.000 --> 00:30:21.130 the RNA/DNA-mediated mechanisms. 00:30:21.130 --> 00:30:24.100 There are DNA-modifying proteins, 00:30:24.100 --> 00:30:26.920 the most important of which is the stuff involved 00:30:26.920 --> 00:30:32.200 in CRISPR-CAS9, which is this relatively new discovery 00:30:32.200 --> 00:30:37.540 about how bacteria are able to use a mechanism that they stole 00:30:37.540 --> 00:30:45.670 from viruses to edit the genetic complement of themselves, 00:30:45.670 --> 00:30:50.110 and more importantly, of other viruses that attack them. 00:30:50.110 --> 00:30:53.120 So it's an antiviral defense mechanism. 00:30:53.120 --> 00:30:56.890 And we're now figuring out how to use it to do gene editing. 00:30:56.890 --> 00:31:01.270 You may have read about this Chinese guy who actually went 00:31:01.270 --> 00:31:05.980 out and edited the genome of a couple of girls who were born 00:31:05.980 --> 00:31:11.050 in China, incorporating some, I think, resistance 00:31:11.050 --> 00:31:14.390 against HIV infections in their genome. 00:31:14.390 --> 00:31:17.080 And of course, this is probably way too early 00:31:17.080 --> 00:31:19.570 to do experiments on human beings, 00:31:19.570 --> 00:31:22.810 because they haven't demonstrated that this is safe. 00:31:22.810 --> 00:31:25.990 But maybe that'll become accepted. 00:31:25.990 --> 00:31:29.680 George Church at Harvard has been going around-- 00:31:29.680 --> 00:31:32.080 he likes to rattle people's chains. 00:31:32.080 --> 00:31:33.850 And he's been going around saying, well, 00:31:33.850 --> 00:31:39.280 the guy, he was unethical and was a slob, but what he's doing 00:31:39.280 --> 00:31:41.110 is a really great idea. 00:31:41.110 --> 00:31:45.480 So we'll see where that goes. 00:31:45.480 --> 00:31:48.750 And then there are these retrotransposons, 00:31:48.750 --> 00:31:53.460 where pieces of DNA in eukarya just pop out 00:31:53.460 --> 00:31:56.670 of wherever they are and insert themselves 00:31:56.670 --> 00:31:59.860 in some other place in the genome. 00:31:59.860 --> 00:32:02.770 And in plants, this happens a lot. 00:32:02.770 --> 00:32:08.550 So for example, wheat seems to have a huge number of copies 00:32:08.550 --> 00:32:13.320 of DNA segments that maybe it had only one of, 00:32:13.320 --> 00:32:16.730 but it's replicated through this mechanism. 00:32:19.690 --> 00:32:23.390 Last bit of complexity-- 00:32:23.390 --> 00:32:26.520 so we have various kinds of RNA. 00:32:26.520 --> 00:32:29.740 There's long non-coding RNA, which 00:32:29.740 --> 00:32:32.640 seems to participate in gene regulation. 00:32:32.640 --> 00:32:39.700 There is RNA interference, that there are these small RNA 00:32:39.700 --> 00:32:44.260 pieces that will actually latch onto the RNA produced 00:32:44.260 --> 00:32:46.770 by the standard genetic mechanism 00:32:46.770 --> 00:32:50.260 and prevent it from being translated into protein. 00:32:50.260 --> 00:32:53.530 This was another Nobel Prize a few years ago. 00:32:53.530 --> 00:32:56.700 Almost everything in this field, if you're first, 00:32:56.700 --> 00:32:58.150 you get a Nobel Prize for it. 00:33:00.960 --> 00:33:03.420 Once the proteins are made, they're 00:33:03.420 --> 00:33:06.010 degraded differentially. 00:33:06.010 --> 00:33:08.130 So there are different mechanisms 00:33:08.130 --> 00:33:11.490 in the cell that destroy certain kinds of proteins 00:33:11.490 --> 00:33:13.660 much faster than others. 00:33:13.660 --> 00:33:16.290 And so the production rate doesn't tell you 00:33:16.290 --> 00:33:19.830 how much is going to be there at any particular time. 00:33:19.830 --> 00:33:24.630 And then there's this secondary and tertiary structure, 00:33:24.630 --> 00:33:27.030 where there's actually-- 00:33:27.030 --> 00:33:27.780 what is it? 00:33:27.780 --> 00:33:31.530 It's a mile of DNA in each of your cells. 00:33:31.530 --> 00:33:34.140 So it wouldn't fit. 00:33:34.140 --> 00:33:41.550 And so it gets wrapped up on these acetylated histones 00:33:41.550 --> 00:33:44.610 to produce something called chromatin. 00:33:44.610 --> 00:33:47.700 And again, we don't quite understand how this all works. 00:33:47.700 --> 00:33:50.850 Because you'd think that if you wrap stuff up like this, 00:33:50.850 --> 00:33:55.290 it would become inaccessible to transcription. 00:33:55.290 --> 00:33:58.560 And therefore, it's not clear how it gets expressed. 00:33:58.560 --> 00:34:01.840 But somehow or other, the cell is able to do that. 00:34:01.840 --> 00:34:06.840 So there's a lot yet to learn in this area. 00:34:06.840 --> 00:34:09.480 Now, the reason we're interested in all this 00:34:09.480 --> 00:34:14.400 is because, if you plot Moore's law for how quickly computers 00:34:14.400 --> 00:34:17.610 are becoming cheaper per performance, 00:34:17.610 --> 00:34:22.500 and you plot the cost of gene sequencing, 00:34:22.500 --> 00:34:24.270 it keeps going down. 00:34:24.270 --> 00:34:29.120 And it goes down much faster even than Moore's law. 00:34:29.120 --> 00:34:31.239 So this is pretty remarkable. 00:34:31.239 --> 00:34:36.070 And it means that, as I said, that $3 dollar first genome now 00:34:36.070 --> 00:34:39.010 costs just a few hundred dollars. 00:34:39.010 --> 00:34:44.420 In fact, if you're just interested in the whole exome, 00:34:44.420 --> 00:34:51.710 so only the 2%, roughly, of the DNA that 00:34:51.710 --> 00:34:55.400 produces genetic coding, you can now 00:34:55.400 --> 00:34:58.490 go to this company, which I have nothing to do with. 00:34:58.490 --> 00:35:00.950 I just pulled this off the web. 00:35:00.950 --> 00:35:09.000 But for $299, they will give you 50-times coverage 00:35:09.000 --> 00:35:12.110 on about six gigabases. 00:35:12.110 --> 00:35:17.300 And if you pay them an extra $100, 00:35:17.300 --> 00:35:19.860 they'll do it at 100x coverage. 00:35:19.860 --> 00:35:22.530 So these techniques are very noisy. 00:35:22.530 --> 00:35:25.340 And so it's important to get lots of replicates 00:35:25.340 --> 00:35:30.430 in order to reassemble what you think is going on. 00:35:30.430 --> 00:35:34.320 A slightly more recent phenomenon is people say, well, 00:35:34.320 --> 00:35:36.930 not only can we sequence your DNA 00:35:36.930 --> 00:35:42.320 but we can sequence the RNA that got transcribed from the DNA. 00:35:42.320 --> 00:35:49.880 And in fact, you can buy a kit for $360 that will take 00:35:49.880 --> 00:35:53.220 the RNA from individual cells-- 00:35:53.220 --> 00:35:58.770 so these are like picoliter amounts of stuff. 00:35:58.770 --> 00:36:04.250 And it will give you the RNA sequence for $360 for up to 100 00:36:04.250 --> 00:36:09.120 cells, so $3, $3.50 per cell. 00:36:09.120 --> 00:36:11.030 So people are very excited. 00:36:11.030 --> 00:36:13.730 And there are now also companies that will 00:36:13.730 --> 00:36:16.580 sell you advanced analysis. 00:36:16.580 --> 00:36:19.100 So they will correlate the data that you 00:36:19.100 --> 00:36:22.310 are getting with different databases 00:36:22.310 --> 00:36:26.540 and figure out whether this represents 00:36:26.540 --> 00:36:30.260 a dominant or a recessive or an x-linked model, 00:36:30.260 --> 00:36:34.760 if you have family familial data and functional annotation 00:36:34.760 --> 00:36:37.250 of candidate genes, et cetera. 00:36:37.250 --> 00:36:40.610 And so, for example, starting about three years ago, 00:36:40.610 --> 00:36:46.070 if you walk into the Dana-Farber with a newly diagnosed cancer, 00:36:46.070 --> 00:36:50.420 a solid-tumor cancer, they will take a sample of that cancer, 00:36:50.420 --> 00:36:54.560 send it off to companies like this, or their own labs, 00:36:54.560 --> 00:36:58.190 and do sequencing and do analysis and try to figure out 00:36:58.190 --> 00:37:02.780 exactly which damaged genes that you have may be causing 00:37:02.780 --> 00:37:07.820 the cancer, and maybe more importantly, since it's still 00:37:07.820 --> 00:37:13.730 a pretty empirical field, which unusual variants of your genes 00:37:13.730 --> 00:37:16.400 suggest that certain drugs are likely to be 00:37:16.400 --> 00:37:20.270 more effective in treating your cancer than other drugs. 00:37:20.270 --> 00:37:25.131 So this has become completely routine in cancer care 00:37:25.131 --> 00:37:27.185 and in a few other domains. 00:37:30.530 --> 00:37:35.250 So now I'm going to switch to a more technical set of material. 00:37:35.250 --> 00:37:38.390 So if you want to characterize disease subtypes 00:37:38.390 --> 00:37:42.710 using gene expression arrays, microarrays, here's 00:37:42.710 --> 00:37:43.740 one way to do it. 00:37:43.740 --> 00:37:47.070 And this is a famous paper by Alizadeh. 00:37:47.070 --> 00:37:51.500 It was essentially the first of this class of papers 00:37:51.500 --> 00:37:54.220 back in 2001, I think. 00:37:54.220 --> 00:37:56.420 Yeah, 2001. 00:37:56.420 --> 00:37:59.840 And since then, there have been probably tens or hundreds 00:37:59.840 --> 00:38:02.420 of thousands of other papers published 00:38:02.420 --> 00:38:06.930 doing similar kinds of analyses on other data sets. 00:38:06.930 --> 00:38:09.510 So what they did is they said, OK, we're 00:38:09.510 --> 00:38:16.200 going to extract the coding RNA. 00:38:16.200 --> 00:38:21.510 We're going to create complementary DNA from it. 00:38:21.510 --> 00:38:24.420 We're going to use a technique to amplify that, 00:38:24.420 --> 00:38:27.660 because we're starting with teeny-tiny quantities. 00:38:27.660 --> 00:38:34.380 And then we're going to take a microarray, which is either 00:38:34.380 --> 00:38:38.580 a glass slide with tens or hundreds of thousands 00:38:38.580 --> 00:38:43.200 of spotted bits of DNA on it or it's 00:38:43.200 --> 00:38:46.770 a silicon chip with wells that, again, 00:38:46.770 --> 00:38:51.360 have tens or hundreds of thousands of bits of DNA in it. 00:38:51.360 --> 00:38:54.640 Now, where does that DNA come from? 00:38:54.640 --> 00:38:57.360 Initially, it was just a random collection 00:38:57.360 --> 00:39:02.790 of pieces of genes from the genome. 00:39:02.790 --> 00:39:05.920 Since then, they've gotten somewhat more sophisticated. 00:39:05.920 --> 00:39:12.090 But the idea is that I'm going to take the amplified cDNA, 00:39:12.090 --> 00:39:15.870 I'm going to mark with one of these jellyfish proteins that 00:39:15.870 --> 00:39:18.750 glows under light, and then I'm going 00:39:18.750 --> 00:39:23.370 to flow it over this slide or over this set of wells. 00:39:23.370 --> 00:39:29.410 And the complementary parts of the complementary DNA 00:39:29.410 --> 00:39:35.556 will stick to the samples of DNA that are in this well. 00:39:35.556 --> 00:39:39.260 OK-- stands to reason. 00:39:39.260 --> 00:39:43.040 An alternative is that you take normal tissue as well 00:39:43.040 --> 00:39:47.150 as, say, the cancerous tissue, you mark the normal tissue 00:39:47.150 --> 00:39:51.770 with green fluorescent jellyfish stuff 00:39:51.770 --> 00:39:54.750 and you mark the cancer with red, 00:39:54.750 --> 00:39:57.260 and then you flow both of them in equal amounts 00:39:57.260 --> 00:39:58.550 over the array. 00:39:58.550 --> 00:40:00.500 That lets you measure a ratio. 00:40:00.500 --> 00:40:03.650 And you don't have as much of a calibration problem 00:40:03.650 --> 00:40:07.640 about trying to figure out the exact value. 00:40:07.640 --> 00:40:10.760 And then you cluster these samples by nearness 00:40:10.760 --> 00:40:12.470 in the expression space. 00:40:12.470 --> 00:40:16.670 And you cluster the genes by expression similarity 00:40:16.670 --> 00:40:18.240 across samples. 00:40:18.240 --> 00:40:20.540 So it used to be called bi-clustering. 00:40:20.540 --> 00:40:24.110 And I'll talk in a few minutes about a particular technique 00:40:24.110 --> 00:40:26.370 for doing this. 00:40:26.370 --> 00:40:30.750 So this is a typical microarray experiment. 00:40:30.750 --> 00:40:34.790 The RNA is turned into its complementary DNA, 00:40:34.790 --> 00:40:37.440 flowed over the microarray chip. 00:40:37.440 --> 00:40:39.720 And you get out a bunch of spots that 00:40:39.720 --> 00:40:43.960 are to various degrees of green and red. 00:40:43.960 --> 00:40:48.370 And then you calculate their ratio. 00:40:48.370 --> 00:40:50.620 And then you do this bi-clustering. 00:40:50.620 --> 00:40:53.370 And what you get is a hierarchical clustering 00:40:53.370 --> 00:40:57.540 of genes and a hierarchical clustering, in their case, 00:40:57.540 --> 00:41:01.240 of breast cancer biopsy specimens that express 00:41:01.240 --> 00:41:02.790 these genes in different ways. 00:41:06.090 --> 00:41:09.150 So this was pretty revolutionary, 00:41:09.150 --> 00:41:12.730 because the answers made sense. 00:41:12.730 --> 00:41:16.590 So when they did this on 19 cell lines 00:41:16.590 --> 00:41:22.230 in 65 breast tumor samples and a whole bunch of genes, 00:41:22.230 --> 00:41:26.400 they came up with a clustering that said, hmm, 00:41:26.400 --> 00:41:31.950 it looks like there are some samples that have 00:41:31.950 --> 00:41:34.810 this endothelial cell cluster. 00:41:34.810 --> 00:41:37.230 So it's a particular kind of problem. 00:41:37.230 --> 00:41:42.240 And you could correlate it with pathology 00:41:42.240 --> 00:41:47.280 from the tumor slides and different subclasses. 00:41:47.280 --> 00:41:51.330 And then this is a very typical kind of heat map 00:41:51.330 --> 00:41:53.820 that you see in this type of study. 00:42:00.270 --> 00:42:04.590 Another study from 65 breast carcinoma samples, 00:42:04.590 --> 00:42:08.700 using the gene list that they curated before, 00:42:08.700 --> 00:42:12.540 looks like it clusters the expression levels 00:42:12.540 --> 00:42:14.040 into these five clusters. 00:42:17.210 --> 00:42:18.500 It's a little hard to look at. 00:42:18.500 --> 00:42:21.770 I mean, when I stare at these, it's not obvious to me 00:42:21.770 --> 00:42:25.790 why the mathematics came up with exactly those clusters rather 00:42:25.790 --> 00:42:27.080 than some others. 00:42:27.080 --> 00:42:30.150 But you can see that there is some sense to it. 00:42:30.150 --> 00:42:34.640 So here you see a lot of greens at this end of it 00:42:34.640 --> 00:42:38.460 and not very much at this end, and vise versa. 00:42:38.460 --> 00:42:40.850 So there is some difference between these clusters. 00:42:40.850 --> 00:42:41.450 Yeah? 00:42:41.450 --> 00:42:43.533 AUDIENCE: How did they come up with the gene list? 00:42:43.533 --> 00:42:46.254 And does anyone ever do this kind of cluster analysis 00:42:46.254 --> 00:42:48.195 without coming up with a gene list first? 00:42:48.195 --> 00:42:49.070 PETER SZOLOVITS: Yes. 00:42:49.070 --> 00:42:52.850 So I'm going to talk in a minute about modern gene-wide 00:42:52.850 --> 00:42:55.670 association studies, where basically you 00:42:55.670 --> 00:42:59.630 say, I'm going to look at every gene known to man. 00:42:59.630 --> 00:43:04.760 So they still have a list, but the list is 20,000 or 25,000. 00:43:04.760 --> 00:43:06.770 It's whatever we know about. 00:43:06.770 --> 00:43:09.310 And that's another way of doing it. 00:43:09.310 --> 00:43:15.830 So what was compelling about this work, this group's work, 00:43:15.830 --> 00:43:21.560 is a later analysis showed that these five subtypes actually 00:43:21.560 --> 00:43:26.930 had different survival rates, and at p-equal 0.01 level 00:43:26.930 --> 00:43:29.030 of statistical significance. 00:43:29.030 --> 00:43:31.130 You've seen these survival curves, of course, 00:43:31.130 --> 00:43:33.500 before from David's lecture. 00:43:33.500 --> 00:43:37.340 But this is pretty impressive that doing something 00:43:37.340 --> 00:43:40.340 that had nothing to do with the clinical condition 00:43:40.340 --> 00:43:41.250 of the patient-- 00:43:41.250 --> 00:43:45.570 this is purely based on their gene expression levels-- 00:43:45.570 --> 00:43:48.740 you were able to find clusters that actually 00:43:48.740 --> 00:43:50.780 behave differently, clinically. 00:43:50.780 --> 00:43:53.840 So some of them do better than others. 00:43:53.840 --> 00:43:57.650 So this paper and this approach to work 00:43:57.650 --> 00:44:02.360 set off a huge set of additional work. 00:44:02.360 --> 00:44:06.110 This was, again, back in the Alizadeh paper. 00:44:06.110 --> 00:44:12.350 They did a similar relationship between 96 samples 00:44:12.350 --> 00:44:15.860 of normal and malignant lymphocytes. 00:44:15.860 --> 00:44:20.180 And they get a similar result, where 00:44:20.180 --> 00:44:23.450 the clusters that they identify here 00:44:23.450 --> 00:44:28.820 correspond to sort of well-understood existing 00:44:28.820 --> 00:44:31.150 types of lymphoma. 00:44:31.150 --> 00:44:35.350 So this, again, gives you some confidence 00:44:35.350 --> 00:44:41.080 that what you're extracting from these genetic correlations 00:44:41.080 --> 00:44:45.820 is meaningful in the terms that people who deal with lymphomas 00:44:45.820 --> 00:44:48.530 think about, about the topic. 00:44:48.530 --> 00:44:51.160 But of course, it can give you much more detail. 00:44:51.160 --> 00:44:53.770 Because people's intuitions may not 00:44:53.770 --> 00:44:59.210 be as effective as these large-scale data analyses. 00:44:59.210 --> 00:45:02.410 So to get to your question about generalizing this, 00:45:02.410 --> 00:45:06.050 I mean, here's one way that I look at this. 00:45:06.050 --> 00:45:12.280 If I list all the genes and I list all the phenotypes-- 00:45:12.280 --> 00:45:14.110 now, we're a little more sure of what 00:45:14.110 --> 00:45:16.760 the genes are than of what the phenotypes are. 00:45:16.760 --> 00:45:19.690 So that's an interesting problem. 00:45:19.690 --> 00:45:23.530 Then I can do a bunch of analyses. 00:45:23.530 --> 00:45:27.460 So what is a phenotype? 00:45:27.460 --> 00:45:31.720 Well, it can be a diagnosed disease, like breast cancer. 00:45:31.720 --> 00:45:35.320 Or it can be the type of lymphoma from the two examples 00:45:35.320 --> 00:45:36.880 I've just shown you. 00:45:36.880 --> 00:45:40.000 It can also be a qualitative or a quantitative trait. 00:45:40.000 --> 00:45:41.020 It could be your weight. 00:45:41.020 --> 00:45:42.340 It could be your eye color. 00:45:42.340 --> 00:45:48.790 It could be almost anything that is clinically known about you. 00:45:48.790 --> 00:45:50.860 And it could even be behavior. 00:45:50.860 --> 00:45:58.940 It could be things like, what is your daily output of Twitter 00:45:58.940 --> 00:45:59.440 posts? 00:46:02.240 --> 00:46:04.670 That's a perfectly reasonable trait. 00:46:04.670 --> 00:46:06.830 I don't know if it's genetically predictable. 00:46:06.830 --> 00:46:12.170 But you'll see some surprising things that are. 00:46:12.170 --> 00:46:14.300 So then, how do you analyze this? 00:46:14.300 --> 00:46:19.070 Well, if you start by looking at a particular phenotype 00:46:19.070 --> 00:46:22.070 and say, what genes are associated with this, 00:46:22.070 --> 00:46:25.250 then you're doing what's called a GWAS, or a Gene-Wide 00:46:25.250 --> 00:46:27.260 Association Study. 00:46:27.260 --> 00:46:29.120 So you look for genetic differences 00:46:29.120 --> 00:46:32.810 that correspond to specific phenotypic differences. 00:46:32.810 --> 00:46:35.810 And usually, you're looking at things like single nucleotide 00:46:35.810 --> 00:46:37.640 polymorphisms. 00:46:37.640 --> 00:46:40.850 So this is places where your genome differs 00:46:40.850 --> 00:46:44.030 from the reference genome, the most common genome 00:46:44.030 --> 00:46:47.760 in the human population, at one particular locus. 00:46:47.760 --> 00:46:51.590 So you have a C instead of a G or something one place 00:46:51.590 --> 00:46:53.120 in your genes. 00:46:53.120 --> 00:46:57.710 Copy number variations, there are stretches of DNA 00:46:57.710 --> 00:47:01.010 that have repeats in them. 00:47:01.010 --> 00:47:03.810 And the number of repeats is variable. 00:47:03.810 --> 00:47:06.020 So one of the most famous ones of these 00:47:06.020 --> 00:47:09.330 is the one associated with Huntington's disease. 00:47:09.330 --> 00:47:14.090 It turns out that if you have up to 20-something repeats 00:47:14.090 --> 00:47:17.790 of a certain section of DNA, you're perfectly healthy. 00:47:17.790 --> 00:47:20.220 But if you're above 30 something, 00:47:20.220 --> 00:47:23.760 then you're going to die of Huntington's disease. 00:47:23.760 --> 00:47:26.780 And again, we don't quite understand these mechanisms. 00:47:26.780 --> 00:47:28.790 But these are empirically known. 00:47:28.790 --> 00:47:31.580 So copy number variations are important, 00:47:31.580 --> 00:47:38.030 gene expression levels, which I've talked about a minute ago. 00:47:38.030 --> 00:47:41.150 But the trick here in a GWAS is to look 00:47:41.150 --> 00:47:44.690 at a very wide set of genes rather than 00:47:44.690 --> 00:47:48.260 just a limited set of samples that you 00:47:48.260 --> 00:47:50.300 know you're interested in. 00:47:50.300 --> 00:47:52.650 Now, the other approach is the opposite, 00:47:52.650 --> 00:47:56.600 which is to say, let's look at a particular gene 00:47:56.600 --> 00:47:59.600 and figure out what's it correlated with. 00:47:59.600 --> 00:48:05.000 And so that's called a PheWAS, a Phenome-Wide Association Study. 00:48:05.000 --> 00:48:10.340 And now what you do is you list all the different phenotypes. 00:48:10.340 --> 00:48:14.180 And you say, well, we can do the same kind of analysis 00:48:14.180 --> 00:48:17.570 to say which of them are disproportionately 00:48:17.570 --> 00:48:22.490 present in people who have that genetic variant. 00:48:22.490 --> 00:48:25.220 So here's what a typical GWAS looks like. 00:48:25.220 --> 00:48:29.190 This is called a Manhattan plot, which I think is pretty funny. 00:48:29.190 --> 00:48:33.540 But it does kind of look like the skyline of Manhattan. 00:48:33.540 --> 00:48:37.730 So this is all of your genes laid out in sequence 00:48:37.730 --> 00:48:40.250 along your chromosomes. 00:48:40.250 --> 00:48:45.980 And you take a particular phenotype and you 00:48:45.980 --> 00:48:51.200 say, what is the difference in the ratio of expression 00:48:51.200 --> 00:48:55.250 levels between people who have this disease and people who 00:48:55.250 --> 00:48:57.200 don't have this disease? 00:48:57.200 --> 00:49:01.140 And something like this gene, whatever it is, clearly there 00:49:01.140 --> 00:49:04.670 is an enormous difference in its expression level. 00:49:04.670 --> 00:49:07.760 And so you would be surprised if this gene didn't have something 00:49:07.760 --> 00:49:09.920 to do with the disease. 00:49:09.920 --> 00:49:15.740 And similarly, you can calculate different significance levels. 00:49:15.740 --> 00:49:18.470 You have to do something like a Bonferroni correction, 00:49:18.470 --> 00:49:23.400 because you are testing so many hypotheses simultaneously. 00:49:23.400 --> 00:49:26.210 And so typically, the top of these lines 00:49:26.210 --> 00:49:29.630 is the Bonferroni-corrected threshold. 00:49:29.630 --> 00:49:33.780 And then you say, OK, this guy, this guy, this guy, this guy, 00:49:33.780 --> 00:49:36.420 and this guy come above that threshold. 00:49:36.420 --> 00:49:38.300 So these are good candidate genes 00:49:38.300 --> 00:49:41.840 to think that may be associated with this disease. 00:49:41.840 --> 00:49:46.140 Now, can you go out and start treating people based on that? 00:49:46.140 --> 00:49:48.380 Well, it's probably not a good idea. 00:49:48.380 --> 00:49:51.740 Because there are many reasons why this analysis 00:49:51.740 --> 00:49:52.640 might have failed. 00:49:52.640 --> 00:49:56.240 All the lessons that you've heard about confounders 00:49:56.240 --> 00:49:58.520 come in very strongly here. 00:49:58.520 --> 00:50:00.620 And so typically, what biologists 00:50:00.620 --> 00:50:03.230 do is they do this kind of analysis. 00:50:03.230 --> 00:50:08.030 They then create a strain of knock-out mice 00:50:08.030 --> 00:50:10.895 who have some analog of whatever disease 00:50:10.895 --> 00:50:12.020 it is that you're studying. 00:50:15.410 --> 00:50:17.870 And they see whether, in fact, knocking out 00:50:17.870 --> 00:50:24.590 a certain gene, like this guy, cures or creates 00:50:24.590 --> 00:50:27.470 the disease that you're interested in in this mouse 00:50:27.470 --> 00:50:28.280 model. 00:50:28.280 --> 00:50:31.130 And then you have a more mechanistic explanation 00:50:31.130 --> 00:50:32.960 for what the relationship might be. 00:50:38.770 --> 00:50:42.010 So basically, you're looking at the ratio 00:50:42.010 --> 00:50:47.790 of the odds of having the disease if you have a SNP, 00:50:47.790 --> 00:50:52.570 or if you have a genetic variant, to having the disease 00:50:52.570 --> 00:50:54.600 if you don't have the genetic variant. 00:50:54.600 --> 00:50:55.210 Yeah? 00:50:55.210 --> 00:50:57.145 AUDIENCE: I'm just curious on the class size. 00:50:57.145 --> 00:50:58.770 It seems like the Bonferroni correction 00:50:58.770 --> 00:51:01.920 is being very limiting here, potentially conservative. 00:51:01.920 --> 00:51:04.930 And I'm curious if there are specific computational 00:51:04.930 --> 00:51:07.060 techniques adapted to this scenario that 00:51:07.060 --> 00:51:10.133 allow you to sort of mine a bit more effectively than those. 00:51:10.133 --> 00:51:11.050 PETER SZOLOVITS: Yeah. 00:51:11.050 --> 00:51:14.740 So if you talk to the statisticians, who 00:51:14.740 --> 00:51:19.030 are more expert at this than the computer scientists typically, 00:51:19.030 --> 00:51:21.160 they will tell you that Bonferroni 00:51:21.160 --> 00:51:24.940 is a very conservative kind of correction. 00:51:24.940 --> 00:51:28.840 And if you can impose some sort of structure 00:51:28.840 --> 00:51:33.370 on the set of genes that you're testing, then you can cheat. 00:51:33.370 --> 00:51:39.100 And you can say, well, you know, these 75 genes actually 00:51:39.100 --> 00:51:41.420 are all part of the same mechanism. 00:51:41.420 --> 00:51:43.780 And we're really testing the mechanism and not 00:51:43.780 --> 00:51:46.000 the individual gene. 00:51:46.000 --> 00:51:49.240 And therefore, instead of making a Bonferroni correction 00:51:49.240 --> 00:51:53.470 for 75 of these guys, we only have to do it for one. 00:51:53.470 --> 00:51:57.610 And so you can reduce the Bonferroni correction that way. 00:51:57.610 --> 00:52:00.490 But people get nervous when you do that. 00:52:00.490 --> 00:52:06.040 Because your incentive as a researcher 00:52:06.040 --> 00:52:10.090 is to show statistically significant results. 00:52:10.090 --> 00:52:12.790 But that whole question of p-values 00:52:12.790 --> 00:52:15.880 keeps coming under discussion. 00:52:15.880 --> 00:52:20.920 So the head of the American Statistical Association, 00:52:20.920 --> 00:52:24.070 about 15 years ago-- he's the Stanford professor. 00:52:24.070 --> 00:52:31.360 And he published what became a very notorious article saying, 00:52:31.360 --> 00:52:33.640 you know, we got it all wrong. 00:52:33.640 --> 00:52:37.505 Statistical significance is not significance 00:52:37.505 --> 00:52:40.450 in the standard English sense of the word. 00:52:40.450 --> 00:52:42.940 And so he called for various other ways 00:52:42.940 --> 00:52:46.840 and was more sympathetic to Bayesian kinds of reasoning 00:52:46.840 --> 00:52:48.250 and things like that. 00:52:48.250 --> 00:52:50.870 So there may be some gradual movement to that. 00:52:50.870 --> 00:52:54.310 But this is a huge can of worms to which we don't have 00:52:54.310 --> 00:52:55.930 a very good mechanistic answer. 00:52:58.660 --> 00:52:59.860 All right. 00:52:59.860 --> 00:53:03.760 So if you do these GWASs-- 00:53:03.760 --> 00:53:06.400 and this is the real problem with them 00:53:06.400 --> 00:53:10.160 is that most of what you see is down here. 00:53:10.160 --> 00:53:15.365 So you have things with common variants. 00:53:18.300 --> 00:53:21.270 But they have very small effect sizes 00:53:21.270 --> 00:53:24.120 when you look at what their effect is 00:53:24.120 --> 00:53:26.590 on a particular disease. 00:53:26.590 --> 00:53:31.540 And so that same Zach Kohane that I mentioned earlier 00:53:31.540 --> 00:53:34.560 has always been challenging people doing this kind of work, 00:53:34.560 --> 00:53:36.950 saying, look-- 00:53:36.950 --> 00:53:42.080 for example, we did a GWAS with Kat Liao, 00:53:42.080 --> 00:53:45.890 who was a guest interviewee here when I was lecturing 00:53:45.890 --> 00:53:47.540 earlier in the semester. 00:53:47.540 --> 00:53:48.930 She's a rheumatologist. 00:53:48.930 --> 00:53:51.360 And we did a gene-wide association study. 00:53:51.360 --> 00:53:54.650 We found a bunch of genes that had odds ratios 00:53:54.650 --> 00:53:59.360 of like 1.1 to 1, 1.2 to 1. 00:53:59.360 --> 00:54:01.160 And they're statistically significant. 00:54:01.160 --> 00:54:03.410 Because if you collect enough data, 00:54:03.410 --> 00:54:06.950 everything is statistically significant. 00:54:06.950 --> 00:54:12.440 But are they significant in the other sense of significance? 00:54:12.440 --> 00:54:15.260 Well, so Zach's argument was that if you 00:54:15.260 --> 00:54:18.800 look at something like the odds ratio of lung cancer 00:54:18.800 --> 00:54:21.910 for people who do and don't smoke, 00:54:21.910 --> 00:54:25.750 the odds ratios is eight. 00:54:25.750 --> 00:54:33.100 So when you compare 1.1 to eight, you should be ashamed. 00:54:33.100 --> 00:54:36.310 You're not doing very well in terms of elucidating 00:54:36.310 --> 00:54:38.470 what the effects really are. 00:54:38.470 --> 00:54:41.650 And so Zack actually has argued very strongly 00:54:41.650 --> 00:54:44.110 that rather than focusing all our attention 00:54:44.110 --> 00:54:49.060 on these genetic factors that have very weak relationships, 00:54:49.060 --> 00:54:52.330 we should instead focus more on clinical things that 00:54:52.330 --> 00:54:56.020 often have stronger predictive relationships. 00:54:56.020 --> 00:54:59.480 And some combination, of course, is best. 00:54:59.480 --> 00:55:04.870 Now, it is true that we know a whole bunch of highly penetrant 00:55:04.870 --> 00:55:06.760 Mendelian mutations. 00:55:06.760 --> 00:55:10.660 So these are ones where, one change in your genome, 00:55:10.660 --> 00:55:14.800 and all of a sudden you have some terrible disease. 00:55:14.800 --> 00:55:19.100 And I think when the Genome Project started in the 1990s, 00:55:19.100 --> 00:55:21.580 there was an expectation that we would 00:55:21.580 --> 00:55:24.520 find a whole bunch more things like that 00:55:24.520 --> 00:55:26.860 from knowing the genome. 00:55:26.860 --> 00:55:29.810 And that expectation was dashed. 00:55:29.810 --> 00:55:34.190 Because what we discovered is that our predecessors 00:55:34.190 --> 00:55:36.380 were actually pretty good at recognizing 00:55:36.380 --> 00:55:40.250 those kinds of diseases, from Mendel 00:55:40.250 --> 00:55:42.290 on, with the wrinkled peas. 00:55:42.290 --> 00:55:45.740 If you see a family in which there's a segregation 00:55:45.740 --> 00:55:48.860 pattern where you can see who has the disease 00:55:48.860 --> 00:55:52.880 and who doesn't and what their relationships are, 00:55:52.880 --> 00:55:55.520 you can get a pretty good idea of what 00:55:55.520 --> 00:55:57.770 genes or what genetic variants are 00:55:57.770 --> 00:56:00.300 associated with that disease. 00:56:00.300 --> 00:56:04.650 And it turns out we had found almost all of them. 00:56:04.650 --> 00:56:08.680 And so there weren't a whole lot more that are highly penetrant 00:56:08.680 --> 00:56:10.590 Mendelian mutations. 00:56:10.590 --> 00:56:14.430 And so what we had is mostly these common variants 00:56:14.430 --> 00:56:17.400 with small effects. 00:56:17.400 --> 00:56:20.610 What's really interesting and worth working on 00:56:20.610 --> 00:56:24.180 is these rare variants with small effects. 00:56:24.180 --> 00:56:30.690 So the mystery kid, like the kid whose case I showed you, 00:56:30.690 --> 00:56:33.750 probably has some interesting genetics 00:56:33.750 --> 00:56:39.290 that is quite uncommon, and obviously, for a long time, 00:56:39.290 --> 00:56:41.300 had a small effect. 00:56:41.300 --> 00:56:44.120 But then all of a sudden, something happened. 00:56:44.120 --> 00:56:48.800 And there is this whole field called unknown disease 00:56:48.800 --> 00:56:53.870 diagnosis that says, what do you do when some weirdo walks in 00:56:53.870 --> 00:56:57.530 off the street and you have no idea what's going on? 00:56:57.530 --> 00:57:01.580 And there are now companies-- so I was a judge in a challenge 00:57:01.580 --> 00:57:04.130 about four or five years ago, where 00:57:04.130 --> 00:57:09.890 we took eight kids like this and we genotyped them, 00:57:09.890 --> 00:57:12.440 and we genotyped their parents and their grandparents 00:57:12.440 --> 00:57:13.400 and their siblings. 00:57:13.400 --> 00:57:15.840 And we took all their clinical data. 00:57:15.840 --> 00:57:19.220 This was with the consent of their parents, of course. 00:57:19.220 --> 00:57:22.250 And we made it available as a contest. 00:57:22.250 --> 00:57:24.860 And we had 20-something participants 00:57:24.860 --> 00:57:27.890 from around the world who tried to figure out something 00:57:27.890 --> 00:57:30.950 useful to say about these kids. 00:57:30.950 --> 00:57:33.410 And you go through a pipeline. 00:57:33.410 --> 00:57:35.870 And we did this in two rounds. 00:57:35.870 --> 00:57:39.380 The first round, the pipelines all looked very different. 00:57:39.380 --> 00:57:41.060 And the second round, a couple of years 00:57:41.060 --> 00:57:44.120 later, the pipelines had pretty much converged. 00:57:44.120 --> 00:57:47.030 And I see now that there is a company that did well 00:57:47.030 --> 00:57:51.020 in one of these challenges that now sells this as a service, 00:57:51.020 --> 00:57:54.560 like I showed you before, different company. 00:57:54.560 --> 00:57:58.880 And so you send them the genetic makeup 00:57:58.880 --> 00:58:02.000 of some kid with a weird condition 00:58:02.000 --> 00:58:05.300 and the genetic makeup of their family, 00:58:05.300 --> 00:58:08.270 and it tries to guess which genes 00:58:08.270 --> 00:58:13.580 might be involved in causing the problem that this child has. 00:58:17.550 --> 00:58:19.510 That's not the answer, of course. 00:58:19.510 --> 00:58:24.790 Because that's just a sort of suspicion of a problem. 00:58:24.790 --> 00:58:28.980 And then you have to go out and do real biological work 00:58:28.980 --> 00:58:31.080 to try to reproduce that scenario 00:58:31.080 --> 00:58:33.690 and see what the effects really are. 00:58:33.690 --> 00:58:37.800 But at least in a couple of cases out of those eight, 00:58:37.800 --> 00:58:42.690 those hints have, in fact, led to a much better understanding 00:58:42.690 --> 00:58:45.630 of what caused the problems in these children. 00:58:48.600 --> 00:58:49.875 That was fun, by the way. 00:58:49.875 --> 00:58:53.400 I got my name as an author on one 00:58:53.400 --> 00:58:55.890 of these things that looks like a high-energy physics 00:58:55.890 --> 00:58:56.920 experiment. 00:58:56.920 --> 00:59:01.080 The first two pages of the paper is just the list of authors. 00:59:01.080 --> 00:59:04.245 So it's kind of interesting. 00:59:06.840 --> 00:59:10.770 Now, here's a more recent study, which 00:59:10.770 --> 00:59:14.640 is a gene-wide association of type 2 diabetes. 00:59:14.640 --> 00:59:16.680 It's not quite gene-wide, because they 00:59:16.680 --> 00:59:18.990 didn't study every locus. 00:59:18.990 --> 00:59:22.680 But they studied a hundred loci that 00:59:22.680 --> 00:59:26.860 have been associated with type 2 diabetes in previous studies. 00:59:26.860 --> 00:59:29.850 So of course, if you're not the first person doing 00:59:29.850 --> 00:59:33.120 this kind of work, you can rely on the literature, where 00:59:33.120 --> 00:59:36.820 other people have already come up with some interesting ideas. 00:59:36.820 --> 00:59:39.210 So they wound up selecting 94 type 00:59:39.210 --> 00:59:42.030 2 diabetes-associated variants. 00:59:42.030 --> 00:59:45.810 So these are the glycemic traits, fasting insulin, 00:59:45.810 --> 00:59:51.390 fasting glucose, et cetera; things about your body, 00:59:51.390 --> 00:59:54.120 your body mass index, height, weight, circumference, 00:59:54.120 --> 00:59:57.630 et cetera; lipid levels of various sorts, 00:59:57.630 --> 01:00:00.390 associations with different diseases, 01:00:00.390 --> 01:00:03.300 coronary artery disease, renal function, et cetera. 01:00:07.110 --> 01:00:09.640 And let me come back to this. 01:00:09.640 --> 01:00:11.800 So what they did is they said, OK, 01:00:11.800 --> 01:00:13.710 here's the way we're going to model this. 01:00:13.710 --> 01:00:15.870 We have an association matrix that 01:00:15.870 --> 01:00:20.220 has 47 traits by 94 genetic factors. 01:00:20.220 --> 01:00:22.740 So we make a matrix out of that. 01:00:22.740 --> 01:00:25.230 And then they did something funny. 01:00:25.230 --> 01:00:26.850 So they doubled the traits. 01:00:26.850 --> 01:00:31.830 The technology for matrix factorization 01:00:31.830 --> 01:00:34.890 is called non-negative matrix factorization. 01:00:34.890 --> 01:00:37.320 And since many of those associations 01:00:37.320 --> 01:00:40.410 were negative, what they did is, for each trait 01:00:40.410 --> 01:00:43.380 that had both positive and negative values, 01:00:43.380 --> 01:00:46.050 they duplicated the column. 01:00:46.050 --> 01:00:51.180 They created one column that had positive associations and one 01:00:51.180 --> 01:00:55.260 column that had the negation of the negative associations 01:00:55.260 --> 01:00:56.740 with zeros everywhere else. 01:00:56.740 --> 01:00:59.262 So that's how they dealt with that problem. 01:00:59.262 --> 01:01:00.720 And then they said, OK, we're going 01:01:00.720 --> 01:01:04.920 to apply matrix factorization to factor X 01:01:04.920 --> 01:01:10.090 into two matrices, W and H. And I drew those here on the board. 01:01:10.090 --> 01:01:12.630 So you have one matrix that-- 01:01:12.630 --> 01:01:17.070 well, this is your original 47 by 94 matrix. 01:01:17.070 --> 01:01:21.570 And the question is, can you find two smaller matrices that 01:01:21.570 --> 01:01:26.640 are 47 by K and K by 94, that when you multiply these 01:01:26.640 --> 01:01:29.700 together, you get back some close approximation 01:01:29.700 --> 01:01:31.560 to that matrix. 01:01:31.560 --> 01:01:34.620 Now, if you've been looking at the literature, 01:01:34.620 --> 01:01:37.590 there are all kinds of ideas like auto-encoders. 01:01:37.590 --> 01:01:42.000 And these are all basically the same underlying idea. 01:01:42.000 --> 01:01:45.210 It's an unsupervised method that says, 01:01:45.210 --> 01:01:48.330 can we find interesting patterns in the data 01:01:48.330 --> 01:01:50.820 by doing some kind of dimension reduction? 01:01:50.820 --> 01:01:55.320 And this is one of those methods for doing dimension reduction. 01:01:55.320 --> 01:02:00.120 So what's nice about this one is that when 01:02:00.120 --> 01:02:07.710 they get their W and H, they predict X from that. 01:02:07.710 --> 01:02:10.800 And then they know, of course, what the error is. 01:02:10.800 --> 01:02:15.270 And they say, well, minimizing that error is our objective. 01:02:15.270 --> 01:02:17.820 So that also lets them get at the question of, 01:02:17.820 --> 01:02:19.860 what's the right K? 01:02:19.860 --> 01:02:21.750 And that's an important problem. 01:02:21.750 --> 01:02:23.820 Because normally clustering methods 01:02:23.820 --> 01:02:25.980 like hierarchical clustering, you 01:02:25.980 --> 01:02:28.860 have to specify what the number of clusters 01:02:28.860 --> 01:02:30.310 is that you're looking for. 01:02:30.310 --> 01:02:33.240 And that's hard to do a priori, whereas this technique 01:02:33.240 --> 01:02:38.110 can suggest at least which one fits the data best. 01:02:38.110 --> 01:02:42.450 And so the loss function is some regularized L2 distance 01:02:42.450 --> 01:02:48.450 between the reconstruction, W times H and X, and some penalty 01:02:48.450 --> 01:02:52.470 terms based on the size of W and H 01:02:52.470 --> 01:02:54.960 coupled by these relevance weights that-- 01:02:54.960 --> 01:02:59.000 you can look at the paper, which I think I referred to in here 01:02:59.000 --> 01:03:01.590 and I asked you to read. 01:03:01.590 --> 01:03:04.050 And then they do give sampling and a whole bunch 01:03:04.050 --> 01:03:08.400 of computational tricks to speed up the process. 01:03:08.400 --> 01:03:13.650 So they got about 17,000 people from four different studies. 01:03:13.650 --> 01:03:15.580 They're all of European ancestry. 01:03:15.580 --> 01:03:18.660 So there's the usual generalization problem of, 01:03:18.660 --> 01:03:23.620 how do you apply this to people from other parts of the world? 01:03:23.620 --> 01:03:28.600 And they did individual-level analysis 01:03:28.600 --> 01:03:33.010 of all the individuals with type 2 diabetes from these. 01:03:33.010 --> 01:03:37.060 And the results were that they found five subtypes-- 01:03:37.060 --> 01:03:47.800 again, five-- which were present on 82.3% of iterations. 01:03:47.800 --> 01:03:50.200 By the way, total random aside, there's 01:03:50.200 --> 01:03:54.040 a wonderful video at Caltech of the woman who 01:03:54.040 --> 01:03:58.660 just made the picture of the black hole shadow. 01:03:58.660 --> 01:04:02.410 And she makes arguments very much like this. 01:04:02.410 --> 01:04:05.830 We tried a whole bunch of different ways 01:04:05.830 --> 01:04:08.500 of coming up with this picture. 01:04:08.500 --> 01:04:11.230 And what we decided was true is whatever 01:04:11.230 --> 01:04:14.350 showed up in almost all of the different methods 01:04:14.350 --> 01:04:15.480 of reconstructing it. 01:04:15.480 --> 01:04:18.370 So this is kind of a similar argument. 01:04:18.370 --> 01:04:22.780 And their interpretations, medically, are that one of them 01:04:22.780 --> 01:04:27.000 is involved with variations in the beta cells. 01:04:27.000 --> 01:04:31.240 So these are the cells in your pancreas that make insulin. 01:04:31.240 --> 01:04:35.600 One of them is in variations in proinsulin, 01:04:35.600 --> 01:04:38.630 which is a predecessor of insulin that 01:04:38.630 --> 01:04:40.660 is under different controls. 01:04:40.660 --> 01:04:47.620 And then three others have to do with obesity, bad things 01:04:47.620 --> 01:04:52.135 about your lipid metabolism, and then your liver function. 01:04:54.860 --> 01:04:57.170 And if you look at their results, 01:04:57.170 --> 01:05:03.410 the top spider diagrams, so the way to interpret these 01:05:03.410 --> 01:05:08.720 is that the middle circle, octagon, 01:05:08.720 --> 01:05:14.030 the one in the very middle, is the one with negative data. 01:05:14.030 --> 01:05:17.330 The one in between that and the outside 01:05:17.330 --> 01:05:19.580 is with zero correlation. 01:05:19.580 --> 01:05:22.820 And the outside one is with positive correlation. 01:05:22.820 --> 01:05:25.970 And what you see is that different factors 01:05:25.970 --> 01:05:29.850 have different influences in these different clusters. 01:05:29.850 --> 01:05:31.400 So these are the factors that are 01:05:31.400 --> 01:05:35.330 most informative in figuring out which cluster somebody belongs 01:05:35.330 --> 01:05:36.230 to. 01:05:36.230 --> 01:05:40.660 And they indeed look considerably different. 01:05:40.660 --> 01:05:42.700 I'm not going to have you read this. 01:05:42.700 --> 01:05:44.470 But it'll be in the slides. 01:05:44.470 --> 01:05:46.480 Now, one thing that's interesting-- 01:05:46.480 --> 01:05:50.770 and again, this won't be on the final exam. 01:05:50.770 --> 01:05:52.087 But look at these numbers. 01:05:52.087 --> 01:05:52.795 They're all tiny. 01:05:59.200 --> 01:06:02.450 Some of them are hugely statistically significant. 01:06:02.450 --> 01:06:09.490 So DI, whatever that is, contributes 0.05 units 01:06:09.490 --> 01:06:17.260 to having beta-cell type of this disease at a p-value of 6.6 01:06:17.260 --> 01:06:20.140 times 10 to the minus 37th. 01:06:20.140 --> 01:06:21.740 So it's definitely there. 01:06:21.740 --> 01:06:22.960 It's definitely an effect. 01:06:22.960 --> 01:06:25.920 But it's not a very big effect. 01:06:25.920 --> 01:06:29.520 And what strikes me every time I look at studies 01:06:29.520 --> 01:06:33.840 like this is just how small those effects are, 01:06:33.840 --> 01:06:36.510 whether you're predicting some output 01:06:36.510 --> 01:06:39.090 like the level of insulin in the patient, 01:06:39.090 --> 01:06:42.480 or whether you're predicting something like a category 01:06:42.480 --> 01:06:44.400 membership, as in this table. 01:06:47.130 --> 01:06:51.510 So as I said, PheWAS is a reverse GWAS. 01:06:51.510 --> 01:06:55.650 And the first paper that introduced the terminology 01:06:55.650 --> 01:07:02.270 was by Josh Denny and colleagues at Vanderbilt in 2010. 01:07:02.270 --> 01:07:07.460 And so they did not quite a phenome-wide association. 01:07:07.460 --> 01:07:12.860 But they said, we're going to take 25,000 samples 01:07:12.860 --> 01:07:16.550 from the Vanderbilt biobank, and we're 01:07:16.550 --> 01:07:19.550 going to take the first 6,000 European Americans 01:07:19.550 --> 01:07:23.170 with samples, no other criteria for selection. 01:07:23.170 --> 01:07:24.560 Why European Americans? 01:07:24.560 --> 01:07:28.130 Because all the GWAS data is about European Americans. 01:07:28.130 --> 01:07:31.140 So they wanted to be able to compare to that. 01:07:31.140 --> 01:07:33.230 And then they said, let's pick not 01:07:33.230 --> 01:07:37.920 one SNP but five different SNPs that we're interested in. 01:07:37.920 --> 01:07:39.620 So they picked these, which are known 01:07:39.620 --> 01:07:42.710 to be associated with coronary artery disease 01:07:42.710 --> 01:07:47.870 and carotid artery stenosis, atrial fibrillation, 01:07:47.870 --> 01:07:51.890 multiple sclerosis and lupus, rheumatoid arthritis 01:07:51.890 --> 01:07:53.100 and Crohn's disease. 01:07:53.100 --> 01:07:55.940 So it's a nice grab-bag of interesting disease 01:07:55.940 --> 01:07:58.130 associations. 01:07:58.130 --> 01:08:01.130 And then the hard work they did was 01:08:01.130 --> 01:08:04.160 they went through the tens of thousands 01:08:04.160 --> 01:08:08.690 of different billing codes that were available. 01:08:08.690 --> 01:08:16.010 And they, by hand, clustered them into 744 case groups 01:08:16.010 --> 01:08:19.010 and said, OK, these are the phenotypes 01:08:19.010 --> 01:08:22.130 that we're interested in. 01:08:22.130 --> 01:08:25.100 And that data set, by the way, is still available. 01:08:25.100 --> 01:08:27.200 And it's been used by a lot of other people, 01:08:27.200 --> 01:08:32.310 because nobody wants to repeat that analysis. 01:08:32.310 --> 01:08:34.279 So now what you see is something very 01:08:34.279 --> 01:08:38.689 similar to what you saw in GWAS, except here, what we 01:08:38.689 --> 01:08:41.359 have is the ICD-9 code group. 01:08:41.359 --> 01:08:46.069 I guess by the time this got published, it was up to 1,000. 01:08:46.069 --> 01:08:53.210 And these are the same kinds of odds ratios 01:08:53.210 --> 01:09:00.439 for the genetic expression of those markers. 01:09:00.439 --> 01:09:06.500 And what you find, again, is that this is the p-equal 0.05. 01:09:06.500 --> 01:09:10.130 That's the Bonferroni-corrected version. 01:09:10.130 --> 01:09:13.609 And only multiple sclerosis comes up 01:09:13.609 --> 01:09:17.840 for this particular SNP, which was one of the ones 01:09:17.840 --> 01:09:19.460 that they expected to come up. 01:09:19.460 --> 01:09:23.720 But they were interested to see what else lights up when 01:09:23.720 --> 01:09:25.729 you do this sort of analysis. 01:09:25.729 --> 01:09:30.529 And what they discovered is that malignant neoplasm 01:09:30.529 --> 01:09:34.939 of the rectum, benign digestive tract neoplasms-- 01:09:34.939 --> 01:09:39.290 so there's something going on about cancer that is somehow 01:09:39.290 --> 01:09:42.600 related to this single-nucleotide polymorphism, 01:09:42.600 --> 01:09:45.120 not at a statistically high enough level, 01:09:45.120 --> 01:09:47.390 but it's still kind of intriguing 01:09:47.390 --> 01:09:49.370 that there may be some relationship there. 01:09:49.370 --> 01:09:50.120 Yeah? 01:09:50.120 --> 01:09:52.827 AUDIENCE: So is this data at all public? 01:09:52.827 --> 01:09:54.410 Or is this at one particular hospital? 01:09:54.410 --> 01:09:55.750 Or who has this data? 01:09:55.750 --> 01:09:57.033 Would it be combined? 01:09:57.033 --> 01:09:57.950 PETER SZOLOVITS: Yeah. 01:09:57.950 --> 01:10:01.430 I don't believe that you can get their data unless-- 01:10:01.430 --> 01:10:02.690 I think, if-- 01:10:02.690 --> 01:10:04.670 I mean, they're pretty good about collaborating 01:10:04.670 --> 01:10:05.690 with people. 01:10:05.690 --> 01:10:11.450 So if you're willing to become a volunteer 01:10:11.450 --> 01:10:15.020 employee at Vanderbilt, they could probably take you. 01:10:15.020 --> 01:10:17.330 But I just made that up. 01:10:17.330 --> 01:10:21.410 But every hospital has very strong controls. 01:10:21.410 --> 01:10:25.340 Now, what is available is the NCBI 01:10:25.340 --> 01:10:28.490 has GEO, the Gene Expression Omnibus, which 01:10:28.490 --> 01:10:30.050 has enormous amounts-- 01:10:30.050 --> 01:10:35.940 like, I think, hundreds of billions of sample data. 01:10:35.940 --> 01:10:40.370 But you don't often know exactly what the sample is from. 01:10:40.370 --> 01:10:42.860 So it comes with an accession number 01:10:42.860 --> 01:10:48.620 and an English description of what kind of data it is. 01:10:48.620 --> 01:10:50.840 And there are actually lots of papers 01:10:50.840 --> 01:10:52.580 where people have done natural language 01:10:52.580 --> 01:10:56.270 processing on those English descriptions in order 01:10:56.270 --> 01:10:59.270 to try to figure out what kind of data this is. 01:10:59.270 --> 01:11:01.140 And then they can make use of it. 01:11:01.140 --> 01:11:03.230 So you can be clever. 01:11:03.230 --> 01:11:04.880 And there's a ton of data out there, 01:11:04.880 --> 01:11:09.090 but it's not well-curated data. 01:11:09.090 --> 01:11:11.700 Now, what's interesting is you don't always 01:11:11.700 --> 01:11:12.720 get what you expect. 01:11:12.720 --> 01:11:16.770 So for example, that SNP was selected 01:11:16.770 --> 01:11:19.680 because it's thought to be associated 01:11:19.680 --> 01:11:23.370 with multiple sclerosis and lupus. 01:11:23.370 --> 01:11:28.550 But in reality, the association with lupus is not significant. 01:11:28.550 --> 01:11:32.580 Its p-value of 0.5, which is not very impressive. 01:11:32.580 --> 01:11:37.110 The association with multiple sclerosis is significant. 01:11:37.110 --> 01:11:40.620 And so they found, in this particular study, 01:11:40.620 --> 01:11:46.190 a couple of things that had been expected but didn't work out. 01:11:46.190 --> 01:11:51.770 So for example, this SNP, which was associated with coronary 01:11:51.770 --> 01:11:56.270 artery disease and thought to be associated with this carotid 01:11:56.270 --> 01:12:00.590 plaque deposition in your carotid artery, just isn't. 01:12:00.590 --> 01:12:04.460 p-value of 0.82 is not impressive at all. 01:12:07.630 --> 01:12:09.370 OK, onward. 01:12:09.370 --> 01:12:12.070 So that was done for SNPs. 01:12:12.070 --> 01:12:14.560 Now, a very popular idea today is 01:12:14.560 --> 01:12:18.700 to look at expression levels, partly because of those prices 01:12:18.700 --> 01:12:22.390 I showed you where you can very cheaply get expression 01:12:22.390 --> 01:12:24.850 levels from lots of samples. 01:12:24.850 --> 01:12:28.060 And so there's this whole notion of Expression Quantitative 01:12:28.060 --> 01:12:32.570 Trait Loci, or EQTL, that says, hey, 01:12:32.570 --> 01:12:36.160 instead of working as hard as the Vanderbilt guys did 01:12:36.160 --> 01:12:40.320 to figure out these hundreds of categories of disease, 01:12:40.320 --> 01:12:44.820 let's just take your gene expression levels 01:12:44.820 --> 01:12:49.960 and use those as defining the trait that we're interested in. 01:12:49.960 --> 01:12:52.530 So now we're looking at the relationship 01:12:52.530 --> 01:12:57.210 between your genome and the expression levels. 01:12:57.210 --> 01:12:59.700 And so you might say, well, that ought to be easy. 01:12:59.700 --> 01:13:03.300 Because if the gene is there, it's going to get expressed. 01:13:03.300 --> 01:13:05.460 But of course, that's not telling you 01:13:05.460 --> 01:13:09.330 whether the gene is being activated or repressed 01:13:09.330 --> 01:13:13.680 or enhanced, or whether any of these other complications 01:13:13.680 --> 01:13:16.080 that I talked about earlier are present. 01:13:16.080 --> 01:13:19.230 And so this is an interesting empirical question. 01:13:19.230 --> 01:13:26.190 And so people say, well, maybe a small genetic variation 01:13:26.190 --> 01:13:31.710 will cause different expression levels of some RNA. 01:13:31.710 --> 01:13:33.450 And we can measure these, and then 01:13:33.450 --> 01:13:36.840 use those to do this kind of analysis. 01:13:41.770 --> 01:13:45.580 So differential expression in different populations-- 01:13:45.580 --> 01:13:48.650 there is evidence that, for example, 01:13:48.650 --> 01:13:53.290 if you take 16 people of African descent, 01:13:53.290 --> 01:13:57.970 then 17% of the genes in a small sample 01:13:57.970 --> 01:14:02.380 of 16 people differ in their expression level 01:14:02.380 --> 01:14:05.350 among those individuals; and similarly, 01:14:05.350 --> 01:14:15.550 26% in this Asian population and 17% to 29% in a HapMap sample. 01:14:15.550 --> 01:14:17.980 Of course, some of these differences 01:14:17.980 --> 01:14:22.720 may be because of confounders like environment, 01:14:22.720 --> 01:14:27.670 different tissues, limited correlation of these expression 01:14:27.670 --> 01:14:30.070 levels to disease phenotypes. 01:14:30.070 --> 01:14:32.890 Nevertheless, this type of analysis 01:14:32.890 --> 01:14:38.290 has uncovered relationships between these EQTLs and asthma 01:14:38.290 --> 01:14:40.270 and Crohn's disease. 01:14:40.270 --> 01:14:42.517 So I'll let you read the conclusion of one 01:14:42.517 --> 01:14:43.225 of these studies. 01:14:56.480 --> 01:15:00.700 So this is saying what I said before, that we probably know 01:15:00.700 --> 01:15:03.350 all the Mendelian diseases. 01:15:03.350 --> 01:15:06.580 So the diseases that we're interested in understanding 01:15:06.580 --> 01:15:09.670 better today are the ones that are not Mendelian, 01:15:09.670 --> 01:15:14.530 but they're some complicated combination of effects 01:15:14.530 --> 01:15:17.770 from different genes. 01:15:17.770 --> 01:15:21.770 And that makes it, of course, a much harder problem. 01:15:21.770 --> 01:15:27.090 There is an interesting recent paper-- 01:15:27.090 --> 01:15:28.360 well, not that recent-- 01:15:28.360 --> 01:15:35.560 2005-- that uses Bayesian network technology 01:15:35.560 --> 01:15:37.280 to try to get at this. 01:15:37.280 --> 01:15:40.930 And so they say, well, if you have some quantitative trait 01:15:40.930 --> 01:15:45.400 locus and you treat the RNA expression level 01:15:45.400 --> 01:15:50.365 as this expression quantitative trait locus, and then 01:15:50.365 --> 01:15:55.780 you take C as some complex trait, which might be a disease 01:15:55.780 --> 01:15:58.000 or it might be a proclivity for something, 01:15:58.000 --> 01:16:02.050 or it might be one of Josh Denny's categories 01:16:02.050 --> 01:16:04.840 or whatever, then there are a number 01:16:04.840 --> 01:16:07.810 of different Bayesian network-style models that you 01:16:07.810 --> 01:16:09.190 can build. 01:16:09.190 --> 01:16:14.290 So you can say, ah, the genetic variant 01:16:14.290 --> 01:16:18.820 causes a difference in gene expression, which 01:16:18.820 --> 01:16:21.580 in turn causes the disease. 01:16:21.580 --> 01:16:24.640 Or you could say, hmm, the genetic trait 01:16:24.640 --> 01:16:27.640 causes the disease, which in turn causes 01:16:27.640 --> 01:16:31.840 the observable difference in gene expression. 01:16:31.840 --> 01:16:38.560 Or you can say that the genetic variant causes 01:16:38.560 --> 01:16:43.120 both the expression level and the disease, 01:16:43.120 --> 01:16:45.280 but they're not necessarily coupled. 01:16:45.280 --> 01:16:48.820 So they may be conditionally independent 01:16:48.820 --> 01:16:50.950 given the genetic variant. 01:16:50.950 --> 01:16:53.380 Or you can have more complex issues, 01:16:53.380 --> 01:16:57.100 like you could have the gene causing changes 01:16:57.100 --> 01:17:01.000 in expression level of a whole bunch of different RNA, 01:17:01.000 --> 01:17:05.470 which combined cause some disease. 01:17:05.470 --> 01:17:08.530 Or you can have different genetic changes 01:17:08.530 --> 01:17:12.040 all impacting the expression of some RNA, which 01:17:12.040 --> 01:17:13.870 causes the disease. 01:17:13.870 --> 01:17:17.680 Or-- just wait for it. 01:17:17.680 --> 01:17:20.600 Oops. 01:17:20.600 --> 01:17:26.010 You can have models like this that say, 01:17:26.010 --> 01:17:28.700 we have some environmental contributions 01:17:28.700 --> 01:17:31.220 and a bunch of different genes which 01:17:31.220 --> 01:17:37.550 affect the expression of a bunch of different EQTLs, which 01:17:37.550 --> 01:17:40.640 cause a bunch of clinical traits, which 01:17:40.640 --> 01:17:44.930 cause changes in a bunch of reactive RNA, which 01:17:44.930 --> 01:17:48.530 cause comorbidities. 01:17:48.530 --> 01:17:51.860 So the approach that they take is 01:17:51.860 --> 01:17:57.800 to say, well, we can generate a large set of hypotheses 01:17:57.800 --> 01:18:01.300 like this, and then just calculate 01:18:01.300 --> 01:18:05.750 the likelihood of the data given each of these hypotheses. 01:18:05.750 --> 01:18:09.680 And whichever one assigns the greatest likelihood to the data 01:18:09.680 --> 01:18:12.185 is most likely to be the one that's close to correct. 01:18:15.050 --> 01:18:18.740 So let me just blast through the rest of this quickly. 01:18:18.740 --> 01:18:22.310 Scaling up genome-phenome association studies-- 01:18:22.310 --> 01:18:26.630 the UK Biobank is sort of like this All of Us project. 01:18:26.630 --> 01:18:30.460 But they do make their data available. 01:18:30.460 --> 01:18:34.340 All of Us will, also, but it hasn't been collected yet. 01:18:34.340 --> 01:18:36.455 UK Biobank has about half a million 01:18:36.455 --> 01:18:41.210 de-identified individuals with full exome sequencing, 01:18:41.210 --> 01:18:46.520 although they only have about 10% of what they want now. 01:18:46.520 --> 01:18:51.110 And many of them will have worn 24-hour activity monitors so 01:18:51.110 --> 01:18:53.760 that we have behavioral data. 01:18:53.760 --> 01:18:56.240 Some of them have had repeat measurements. 01:18:56.240 --> 01:18:58.880 They do online questionnaires. 01:18:58.880 --> 01:19:02.930 About a fifth of them will have imaging. 01:19:02.930 --> 01:19:05.730 And it's linked to their electronic health record. 01:19:05.730 --> 01:19:07.490 So we know if they died or if they 01:19:07.490 --> 01:19:12.230 had cancer or various hospital episodes, et cetera. 01:19:12.230 --> 01:19:19.640 And there's a website here which publishes the latest analyses. 01:19:19.640 --> 01:19:23.180 And so you see, on April 18, genetic variants that 01:19:23.180 --> 01:19:27.050 protect against obesity and type 2 diabetes discovered, 01:19:27.050 --> 01:19:30.530 moderate with meat-eaters are at risk of bowel cancer, 01:19:30.530 --> 01:19:33.930 and research identifies genetic causes of poor sleep. 01:19:33.930 --> 01:19:35.690 So this is all over the place. 01:19:35.690 --> 01:19:38.390 But these are all the studies that are being done by this. 01:19:42.050 --> 01:19:42.910 I'll skip this. 01:19:42.910 --> 01:19:46.190 But there's a group here at MGH and the Broad that 01:19:46.190 --> 01:19:49.730 is using this data to do, large-scale, 01:19:49.730 --> 01:19:54.860 many, many gene-wide association studies. 01:19:54.860 --> 01:19:57.830 And one of the things that I promised you, 01:19:57.830 --> 01:20:00.860 which is interesting, is from these studies, they say, 01:20:00.860 --> 01:20:05.370 well, the heritability of height is pretty good. 01:20:05.370 --> 01:20:10.910 It's about 0.46 with a p-value of 10 to the minus 109th. 01:20:10.910 --> 01:20:14.660 So your height is definitely determined, in large part, 01:20:14.660 --> 01:20:16.580 by your parents' height. 01:20:16.580 --> 01:20:18.650 But what's interesting is that whether you 01:20:18.650 --> 01:20:21.680 get a college degree or not is determined 01:20:21.680 --> 01:20:24.500 by whether your parents got a college degree or not. 01:20:24.500 --> 01:20:27.440 This is probably not genetic. 01:20:27.440 --> 01:20:29.540 Or it's only partly genetic. 01:20:29.540 --> 01:20:34.850 But it clearly has confounders us from money and social status 01:20:34.850 --> 01:20:36.770 and various things like that. 01:20:36.770 --> 01:20:40.250 And then what I found amusing is that even 01:20:40.250 --> 01:20:47.000 TV-watching is partly heritable from your genetics. 01:20:50.850 --> 01:20:55.230 Fortunately, my parents watch a lot of TV. 01:20:55.230 --> 01:20:57.150 The last thing I wanted to mention, 01:20:57.150 --> 01:20:59.220 but I'm not going to have time to get into it, 01:20:59.220 --> 01:21:02.390 is this notion of gene set enrichment analysis. 01:21:02.390 --> 01:21:05.490 It's what I was saying before, that genes typically 01:21:05.490 --> 01:21:08.260 don't act by themselves. 01:21:08.260 --> 01:21:11.400 And so if you think back on high school biology, 01:21:11.400 --> 01:21:14.300 you probably learned about the Krebs cycle 01:21:14.300 --> 01:21:17.170 that powers cellular mechanisms. 01:21:17.170 --> 01:21:19.440 So if you break any part of that cycle, 01:21:19.440 --> 01:21:21.940 your cells don't get enough energy. 01:21:21.940 --> 01:21:24.570 And so it stands to reason that if you 01:21:24.570 --> 01:21:27.840 want to understand that sort of metabolism, 01:21:27.840 --> 01:21:30.510 you shouldn't be looking at an individual gene. 01:21:30.510 --> 01:21:33.330 But you should be looking at all of the genes that 01:21:33.330 --> 01:21:36.070 are involved in that process. 01:21:36.070 --> 01:21:39.450 And so there have been many attempts to try to do this. 01:21:39.450 --> 01:21:46.560 The Broad Institute here has a set of, originally, 1,300 01:21:46.560 --> 01:21:48.710 biologically-defined gene sets. 01:21:48.710 --> 01:21:52.350 So these were ones that interacted with each other 01:21:52.350 --> 01:21:55.750 in controlling some important mechanism in the body. 01:21:55.750 --> 01:21:58.380 They're now up to 18,000. 01:21:58.380 --> 01:22:02.190 For example, genes involved in oxidative phosphorylation 01:22:02.190 --> 01:22:05.790 and muscle tissue show reduced expression in diabetics, 01:22:05.790 --> 01:22:09.300 although the average decrease per gene is only 20%. 01:22:09.300 --> 01:22:11.040 So they have these sets. 01:22:11.040 --> 01:22:15.060 And from those, there is a very nice technique 01:22:15.060 --> 01:22:18.090 that is able to pull-- 01:22:24.300 --> 01:22:27.180 it's essentially a way of strengthening 01:22:27.180 --> 01:22:31.710 the gene-wide associations by allowing you to associate them 01:22:31.710 --> 01:22:33.750 with these sets of genes. 01:22:33.750 --> 01:22:36.810 And the approach that they take is quite clever. 01:22:36.810 --> 01:22:40.930 They say, if we take all the genes in a gene set 01:22:40.930 --> 01:22:45.430 and we order them by their correlation with whatever trait 01:22:45.430 --> 01:22:49.450 we're interested in, then the genes 01:22:49.450 --> 01:22:52.420 that are closer to the beginning of that 01:22:52.420 --> 01:22:54.370 are more likely to be involved. 01:22:54.370 --> 01:22:57.970 Because they're the ones that are most strongly associated. 01:22:57.970 --> 01:23:00.790 And so they have this random walk process 01:23:00.790 --> 01:23:04.060 that find sort of the maximum place where 01:23:04.060 --> 01:23:06.640 you can say anything before that is 01:23:06.640 --> 01:23:08.890 likely to be associated with the disease 01:23:08.890 --> 01:23:11.080 that you're interested in. 01:23:11.080 --> 01:23:16.930 And they've had a number of successes of showing enrichment 01:23:16.930 --> 01:23:22.870 in various diseases and various biological factors. 01:23:22.870 --> 01:23:26.650 The last thing I want to say is a little bit disappointing. 01:23:26.650 --> 01:23:30.310 I was just really looking for the killer paper 01:23:30.310 --> 01:23:32.890 to talk about that uses some really 01:23:32.890 --> 01:23:36.160 sophisticated deep learning, machine learning. 01:23:36.160 --> 01:23:40.370 And as far as I can tell, it doesn't exist yet. 01:23:40.370 --> 01:23:45.700 So most of these methods are based on clustering techniques 01:23:45.700 --> 01:23:49.000 on clever ideas, like the one for gene 01:23:49.000 --> 01:23:52.270 set enrichment analysis. 01:23:52.270 --> 01:23:55.780 But they're not neural network types of techniques. 01:23:55.780 --> 01:23:58.540 They're not immensely sophisticated. 01:23:58.540 --> 01:24:02.080 So what you see coming up is things like Bayesian networks 01:24:02.080 --> 01:24:06.040 and clustering and matrix factorization and so on, which 01:24:06.040 --> 01:24:10.390 sort of sound like 10-, 15-, 20-year-old technologies. 01:24:10.390 --> 01:24:15.770 And I haven't seen examples yet of the hot off the presses, 01:24:15.770 --> 01:24:19.870 we built a 83-layer neural network 01:24:19.870 --> 01:24:23.110 that outperforms these other methods. 01:24:23.110 --> 01:24:25.180 I suspect that that's coming. 01:24:25.180 --> 01:24:28.630 It just hasn't hit yet, as far as I know. 01:24:28.630 --> 01:24:31.760 If you know of such papers, by all means, let me know. 01:24:31.760 --> 01:24:32.260 All right. 01:24:32.260 --> 01:24:34.050 Thank you.