WEBVTT 00:00:00.000 --> 00:00:00.053 align:middle line:90% 00:00:00.053 --> 00:00:01.470 align:middle line:84% SCOTT WEISS: So this is an outline 00:00:01.470 --> 00:00:03.220 align:middle line:84% of what I'm going to talk about, and we're 00:00:03.220 --> 00:00:06.240 align:middle line:84% going to begin by of getting at this question of why 00:00:06.240 --> 00:00:10.270 align:middle line:84% complex trait human genetics is so difficult. 00:00:10.270 --> 00:00:12.480 align:middle line:84% And then go through each of the steps 00:00:12.480 --> 00:00:15.900 align:middle line:84% that you would do if you were actually doing this work. 00:00:15.900 --> 00:00:18.150 align:middle line:84% The first question you would get asked on an NIH grant 00:00:18.150 --> 00:00:21.060 align:middle line:84% application is your phenotype heritable? 00:00:21.060 --> 00:00:24.090 align:middle line:84% And so some evidence of heritability, or doing a study 00:00:24.090 --> 00:00:27.240 align:middle line:84% to determine heritability, is often the first step 00:00:27.240 --> 00:00:29.353 align:middle line:90% in a genetic study. 00:00:29.353 --> 00:00:31.770 align:middle line:84% And so we're going to talk a little bit about heritability 00:00:31.770 --> 00:00:34.530 align:middle line:84% and how you define that, and then identifying disease 00:00:34.530 --> 00:00:38.490 align:middle line:84% phenotypes-- so that's a critical issue-- 00:00:38.490 --> 00:00:40.470 align:middle line:84% the difference between a subphenotype 00:00:40.470 --> 00:00:46.055 align:middle line:84% and an intermediate phenotype, how 00:00:46.055 --> 00:00:47.430 align:middle line:84% you want to look at those things, 00:00:47.430 --> 00:00:50.280 align:middle line:84% then developing your study design. 00:00:50.280 --> 00:00:52.920 align:middle line:84% And the paper that I just was telling you about, 00:00:52.920 --> 00:00:55.230 align:middle line:84% this paper that's coming out in science, 00:00:55.230 --> 00:00:58.380 align:middle line:84% these people looked at two relative genetic isolates-- 00:00:58.380 --> 00:01:02.460 align:middle line:84% the population of Finland and the Saguenay-Lac-Saint-Jean 00:01:02.460 --> 00:01:06.480 align:middle line:84% population in Northeast Quebec, French Canadians. 00:01:06.480 --> 00:01:08.580 align:middle line:84% So these are populations where they 00:01:08.580 --> 00:01:11.020 align:middle line:84% had a limited number of founders, 00:01:11.020 --> 00:01:14.310 align:middle line:84% and so the thought is that they're more 00:01:14.310 --> 00:01:16.530 align:middle line:84% genetically homogeneous, and it might be easier 00:01:16.530 --> 00:01:19.200 align:middle line:84% to find genes in these populations. 00:01:19.200 --> 00:01:24.000 align:middle line:84% But what's the big concern, if you found a gene in the Finns? 00:01:24.000 --> 00:01:26.617 align:middle line:84% What would be your biggest concern about that? 00:01:26.617 --> 00:01:28.200 align:middle line:84% AUDIENCE: That it would be replicable? 00:01:28.200 --> 00:01:29.075 align:middle line:90% SCOTT WEISS: Exactly. 00:01:29.075 --> 00:01:32.460 align:middle line:84% Maybe that gene might not replicate or be 00:01:32.460 --> 00:01:35.440 align:middle line:84% a significant gene in an outbred population 00:01:35.440 --> 00:01:37.560 align:middle line:84% in a country such as the United States, where 00:01:37.560 --> 00:01:42.120 align:middle line:84% there's a lot of ethnic variation and diversity. 00:01:42.120 --> 00:01:46.830 align:middle line:84% So it may be easier to find genes in genetic isolates, 00:01:46.830 --> 00:01:50.370 align:middle line:84% but it may be more difficult in terms 00:01:50.370 --> 00:01:51.810 align:middle line:90% of their generalizability. 00:01:51.810 --> 00:01:57.210 align:middle line:84% So give me some other examples of relative genetic isolates 00:01:57.210 --> 00:01:58.230 align:middle line:90% around the world-- 00:01:58.230 --> 00:02:00.420 align:middle line:84% populations that would be considered 00:02:00.420 --> 00:02:01.920 align:middle line:90% relative genetic isolates? 00:02:01.920 --> 00:02:03.390 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:02:03.390 --> 00:02:04.140 align:middle line:90% AUDIENCE: Iceland. 00:02:04.140 --> 00:02:04.590 align:middle line:90% SCOTT WEISS: Iceland. 00:02:04.590 --> 00:02:05.790 align:middle line:90% Yes, absolutely, Iceland. 00:02:05.790 --> 00:02:08.520 align:middle line:84% So that's where Decode is working-- 00:02:08.520 --> 00:02:11.820 align:middle line:84% Decode is our number one competitor in COPD research. 00:02:11.820 --> 00:02:18.340 align:middle line:84% They're the only company that's actually doing COPD stuff. 00:02:18.340 --> 00:02:19.080 align:middle line:90% What else? 00:02:19.080 --> 00:02:20.768 align:middle line:90% Where else could you go? 00:02:20.768 --> 00:02:21.768 align:middle line:90% AUDIENCE: South America. 00:02:21.768 --> 00:02:25.420 align:middle line:84% [INAUDIBLE] generally from Iceland [INAUDIBLE] 00:02:25.420 --> 00:02:27.930 align:middle line:90% in South America. 00:02:27.930 --> 00:02:35.290 align:middle line:84% I guess tribal populations in South America and [INAUDIBLE].. 00:02:35.290 --> 00:02:37.480 align:middle line:84% SCOTT WEISS: Maybe-- what might be 00:02:37.480 --> 00:02:40.900 align:middle line:84% the disadvantage of a tribal population, 00:02:40.900 --> 00:02:45.520 align:middle line:84% or disadvantage of a small island population 00:02:45.520 --> 00:02:49.240 align:middle line:84% like Tristan da Cunha, which is where they first went to do 00:02:49.240 --> 00:02:52.350 align:middle line:90% genetic isolate work in asthma? 00:02:52.350 --> 00:02:53.950 align:middle line:84% AUDIENCE: The small number of people. 00:02:53.950 --> 00:02:55.700 align:middle line:84% SCOTT WEISS: Yeah, small number of people. 00:02:55.700 --> 00:02:59.020 align:middle line:84% So you've got a limited number of meiosis 00:02:59.020 --> 00:03:00.220 align:middle line:90% in a population like that. 00:03:00.220 --> 00:03:03.050 align:middle line:84% You're not going to get too many combinations. 00:03:03.050 --> 00:03:05.440 align:middle line:84% So you just run out of gas because you don't 00:03:05.440 --> 00:03:07.030 align:middle line:90% have a big enough sample size. 00:03:07.030 --> 00:03:08.610 align:middle line:90% AUDIENCE: Swiss. 00:03:08.610 --> 00:03:10.480 align:middle line:84% Swiss families up in the Alps somewhere. 00:03:10.480 --> 00:03:12.250 align:middle line:90% SCOTT WEISS: Possibly. 00:03:12.250 --> 00:03:18.520 align:middle line:84% The Finns, the Swiss, the northern Netherlands, 00:03:18.520 --> 00:03:24.987 align:middle line:90% the Ashkenazi Jews, Costa Rica. 00:03:24.987 --> 00:03:25.570 align:middle line:90% AUDIENCE: Why? 00:03:25.570 --> 00:03:27.070 align:middle line:90% Why Costa Rica? 00:03:27.070 --> 00:03:29.530 align:middle line:90% SCOTT WEISS: Why Costa Rica? 00:03:29.530 --> 00:03:33.070 align:middle line:84% Because Costa Rica is surrounded by volcanoes. 00:03:33.070 --> 00:03:34.990 align:middle line:84% It was the one place in Central America 00:03:34.990 --> 00:03:40.210 align:middle line:84% that the Spanish-- there's not a huge Spanish influx 00:03:40.210 --> 00:03:41.930 align:middle line:90% because there was no gold there. 00:03:41.930 --> 00:03:44.530 align:middle line:84% So there's a limited number of Spanish founders 00:03:44.530 --> 00:03:46.220 align:middle line:90% in the 15th century. 00:03:46.220 --> 00:03:50.380 align:middle line:84% They pushed the Indians to the periphery, 00:03:50.380 --> 00:03:53.260 align:middle line:84% and they settled the Central Valley of Costa Rica. 00:03:53.260 --> 00:03:58.330 align:middle line:84% So you've got this 200 founders in the 14th century, 00:03:58.330 --> 00:04:02.230 align:middle line:84% very little intermarriage, and perfect church records-- 00:04:02.230 --> 00:04:04.750 align:middle line:90% very large pedigrees. 00:04:04.750 --> 00:04:08.620 align:middle line:84% It's probably the closest, next to the Saguenay--Lac-Saint-Jean 00:04:08.620 --> 00:04:13.450 align:middle line:84% population in Quebec, genetic isolates in the Western 00:04:13.450 --> 00:04:13.968 align:middle line:90% hemisphere. 00:04:13.968 --> 00:04:15.760 align:middle line:84% And we're actually doing a big study there. 00:04:15.760 --> 00:04:19.089 align:middle line:84% We actually have six big pedigrees 00:04:19.089 --> 00:04:23.140 align:middle line:84% with over 120 people in each pedigree. 00:04:23.140 --> 00:04:25.450 align:middle line:84% We've just finished the collection of these pedigrees. 00:04:25.450 --> 00:04:28.810 align:middle line:84% We're about to do a genome scan that Tom is actually 00:04:28.810 --> 00:04:29.620 align:middle line:90% doing for us. 00:04:29.620 --> 00:04:32.230 align:middle line:84% AUDIENCE: [INAUDIBLE] most of the people 00:04:32.230 --> 00:04:34.270 align:middle line:84% are isolated by geography, as you've 00:04:34.270 --> 00:04:37.360 align:middle line:84% mentioned with the exception of Ashkenazi Jews. 00:04:37.360 --> 00:04:46.480 align:middle line:84% So to broaden, I guess, the definition of genetic isolate 00:04:46.480 --> 00:04:49.450 align:middle line:84% to include things like economic-- like, things 00:04:49.450 --> 00:04:50.950 align:middle line:84% that usually [? happen to ?] people, 00:04:50.950 --> 00:04:54.610 align:middle line:84% like royal families back in the old days. 00:04:54.610 --> 00:04:57.940 align:middle line:84% They only intermarried because of social status [INAUDIBLE].. 00:04:57.940 --> 00:05:01.360 align:middle line:84% But now there's things where people with certain economic 00:05:01.360 --> 00:05:06.010 align:middle line:84% status [INAUDIBLE] in the coming years [INAUDIBLE] see a certain 00:05:06.010 --> 00:05:08.158 align:middle line:84% set of people where they marry and-- 00:05:08.158 --> 00:05:10.450 align:middle line:84% SCOTT WEISS: Yeah, it turns out, I think, the geography 00:05:10.450 --> 00:05:11.590 align:middle line:90% is actually a much better-- 00:05:11.590 --> 00:05:18.785 align:middle line:90% 00:05:18.785 --> 00:05:19.955 align:middle line:90% [INTERPOSING VOICES] 00:05:19.955 --> 00:05:21.580 align:middle line:84% AUDIENCE: --happening to royal families 00:05:21.580 --> 00:05:24.940 align:middle line:84% where people have a chance to procreate [INAUDIBLE] proximity 00:05:24.940 --> 00:05:27.850 align:middle line:90% as well? 00:05:27.850 --> 00:05:31.960 align:middle line:84% SCOTT WEISS: I think the geography is a lot better. 00:05:31.960 --> 00:05:36.250 align:middle line:84% Historically, it's a much more reliable guide 00:05:36.250 --> 00:05:38.680 align:middle line:84% to a genetic isolate than any social convention. 00:05:38.680 --> 00:05:43.315 align:middle line:84% I think, exactly for the reasons that Zach-- 00:05:43.315 --> 00:05:45.610 align:middle line:84% AUDIENCE: Aren't groups like the Ashkenazi Jews 00:05:45.610 --> 00:05:47.650 align:middle line:84% AUDIENCE: There's so strong social pressure. 00:05:47.650 --> 00:05:49.750 align:middle line:84% AUDIENCE: Yeah, like they [INAUDIBLE].. 00:05:49.750 --> 00:05:52.010 align:middle line:84% SCOTT WEISS: They may be the exception to the rule, 00:05:52.010 --> 00:05:56.740 align:middle line:84% but I think populations like Iceland, Costa Rica, Tristan da 00:05:56.740 --> 00:05:58.120 align:middle line:90% Cunha, Finland-- 00:05:58.120 --> 00:05:59.780 align:middle line:90% AUDIENCE: How about the Amish? 00:05:59.780 --> 00:06:04.030 align:middle line:84% SCOTT WEISS: The Amish would fall into that group. 00:06:04.030 --> 00:06:07.840 align:middle line:84% The Mormons-- maybe a little bit less so. 00:06:07.840 --> 00:06:10.000 align:middle line:84% The big advantage of the Mormons is not so much 00:06:10.000 --> 00:06:11.710 align:middle line:84% that they're a relative genetic isolate, 00:06:11.710 --> 00:06:14.290 align:middle line:84% but that they have very large families 00:06:14.290 --> 00:06:16.210 align:middle line:84% and they have very good church records. 00:06:16.210 --> 00:06:20.470 align:middle line:84% So those are other characteristics 00:06:20.470 --> 00:06:22.120 align:middle line:90% that are helpful. 00:06:22.120 --> 00:06:25.120 align:middle line:84% But the idea behind the genetic isolate 00:06:25.120 --> 00:06:31.240 align:middle line:84% is that you've got a relatively homogeneous set of alleles 00:06:31.240 --> 00:06:33.100 align:middle line:84% that's circulating in the population. 00:06:33.100 --> 00:06:38.290 align:middle line:84% Mormons do intermarry, and so there may be-- 00:06:38.290 --> 00:06:41.290 align:middle line:84% I don't know that you'd consider them really a genetic isolate, 00:06:41.290 --> 00:06:45.670 align:middle line:84% whereas I think that the Amish and the Hutterites 00:06:45.670 --> 00:06:50.590 align:middle line:84% and the Ashkenazi-- these are people where there 00:06:50.590 --> 00:06:53.080 align:middle line:84% is some set of code social conventions 00:06:53.080 --> 00:06:56.320 align:middle line:84% that those people are much more likely to intermarry. 00:06:56.320 --> 00:06:58.420 align:middle line:84% AUDIENCE: The reason I ask is because it seems 00:06:58.420 --> 00:07:04.713 align:middle line:84% like the trend might be, especially nowadays 00:07:04.713 --> 00:07:06.130 align:middle line:84% with transportation the way it is, 00:07:06.130 --> 00:07:09.130 align:middle line:84% that finding geographic isolated populations 00:07:09.130 --> 00:07:12.100 align:middle line:84% is going to decline and [INAUDIBLE].. 00:07:12.100 --> 00:07:16.450 align:middle line:84% SCOTT WEISS: This isn't the only way to do this. 00:07:16.450 --> 00:07:18.340 align:middle line:84% And I think it is important to make 00:07:18.340 --> 00:07:25.660 align:middle line:84% the distinction between linkage and fine mapping, 00:07:25.660 --> 00:07:30.350 align:middle line:84% and those two things may be somewhat different. 00:07:30.350 --> 00:07:33.250 align:middle line:84% The advantages of an outbred population 00:07:33.250 --> 00:07:36.670 align:middle line:84% is just that the degree of linkage disequilibrium 00:07:36.670 --> 00:07:37.900 align:middle line:90% will be relatively narrow. 00:07:37.900 --> 00:07:40.810 align:middle line:84% In some of these genetic isolated populations, 00:07:40.810 --> 00:07:42.490 align:middle line:84% the degree of linkage disequilibrium 00:07:42.490 --> 00:07:43.660 align:middle line:90% can be very large. 00:07:43.660 --> 00:07:47.920 align:middle line:84% That means you can do the linkage part pretty 00:07:47.920 --> 00:07:50.890 align:middle line:84% effectively, but the association part is more difficult 00:07:50.890 --> 00:07:53.740 align:middle line:84% because you've got these big LB blocks that you've 00:07:53.740 --> 00:08:00.795 align:middle line:84% got to work with, and you may not be able to get to the gene. 00:08:00.795 --> 00:08:02.420 align:middle line:84% So anyway, we'll go through this stuff. 00:08:02.420 --> 00:08:05.480 align:middle line:90% 00:08:05.480 --> 00:08:07.210 align:middle line:84% These are all of the steps that you 00:08:07.210 --> 00:08:09.250 align:middle line:84% would do in a typical study, and we'll just 00:08:09.250 --> 00:08:10.550 align:middle line:90% go through each of these. 00:08:10.550 --> 00:08:12.247 align:middle line:84% So I'm going to use asthma as my example 00:08:12.247 --> 00:08:14.080 align:middle line:84% because this is the disease I know the best, 00:08:14.080 --> 00:08:17.170 align:middle line:84% and the point here is that asthma prevalence in Western 00:08:17.170 --> 00:08:19.580 align:middle line:84% developed countries has gone up a lot. 00:08:19.580 --> 00:08:21.430 align:middle line:84% So do you think that this is a genetic thing 00:08:21.430 --> 00:08:24.170 align:middle line:84% or do you think it's something else? 00:08:24.170 --> 00:08:28.780 align:middle line:84% So over the 20-year period, '80 to 2000, 00:08:28.780 --> 00:08:31.610 align:middle line:84% we've had more than a doubling in the number of cases. 00:08:31.610 --> 00:08:33.850 align:middle line:84% So over a 20-year period, doubling 00:08:33.850 --> 00:08:35.830 align:middle line:90% in the number of cases? 00:08:35.830 --> 00:08:36.580 align:middle line:90% AUDIENCE: Genetic? 00:08:36.580 --> 00:08:38.213 align:middle line:84% SCOTT WEISS: Genetic or environment? 00:08:38.213 --> 00:08:39.130 align:middle line:90% AUDIENCE: Environment. 00:08:39.130 --> 00:08:41.788 align:middle line:90% SCOTT WEISS: Why? 00:08:41.788 --> 00:08:44.358 align:middle line:84% AUDIENCE: Environment [INAUDIBLE].. 00:08:44.358 --> 00:08:46.150 align:middle line:84% SCOTT WEISS: So there are three potential-- 00:08:46.150 --> 00:08:50.260 align:middle line:84% what are the three population genetic mechanisms that 00:08:50.260 --> 00:08:54.400 align:middle line:84% something like this could occur-- 00:08:54.400 --> 00:08:57.190 align:middle line:90% a genetic explanation? 00:08:57.190 --> 00:09:02.320 align:middle line:84% So I'll give you at least half credit, 00:09:02.320 --> 00:09:05.920 align:middle line:84% but you're definitely not 100% right. 00:09:05.920 --> 00:09:10.540 align:middle line:84% The first genetic mechanism would be spontaneous mutation. 00:09:10.540 --> 00:09:13.240 align:middle line:84% You had some spontaneous mutation, 00:09:13.240 --> 00:09:17.230 align:middle line:84% and it caused this epidemic of diseases. 00:09:17.230 --> 00:09:20.270 align:middle line:90% Is that possible? 00:09:20.270 --> 00:09:22.850 align:middle line:84% Well, you already said no, it's not possible. 00:09:22.850 --> 00:09:24.650 align:middle line:84% And you're right, it's not, but you 00:09:24.650 --> 00:09:27.290 align:middle line:84% have to know what the spontaneous mutation rate is, 00:09:27.290 --> 00:09:31.130 align:middle line:84% which is about 1 times 10 to the minus 8 base pairs 00:09:31.130 --> 00:09:32.600 align:middle line:90% per generation. 00:09:32.600 --> 00:09:34.700 align:middle line:90% So it's pretty low. 00:09:34.700 --> 00:09:38.030 align:middle line:84% We're spontaneously mutating all of the time, 00:09:38.030 --> 00:09:40.880 align:middle line:84% but we're not spontaneously mutating fast enough 00:09:40.880 --> 00:09:44.370 align:middle line:84% to double the number of cases in a 20-year period of time. 00:09:44.370 --> 00:09:50.210 align:middle line:84% So what's the second possible genetic mechanism? 00:09:50.210 --> 00:09:52.490 align:middle line:90% Natural selection, right? 00:09:52.490 --> 00:09:57.050 align:middle line:84% So is natural selection going to do-- 00:09:57.050 --> 00:09:58.468 align:middle line:90% AUDIENCE: [INAUDIBLE]. 00:09:58.468 --> 00:10:01.010 align:middle line:84% SCOTT WEISS: Well, particularly with the disease like asthma, 00:10:01.010 --> 00:10:03.110 align:middle line:84% where there's no selection pressure 00:10:03.110 --> 00:10:07.460 align:middle line:84% and no reproductive advantages and disadvantages-- 00:10:07.460 --> 00:10:09.710 align:middle line:84% you all know plenty of people with asthma, 00:10:09.710 --> 00:10:12.420 align:middle line:84% and they're able to reproduce just like everybody else. 00:10:12.420 --> 00:10:14.900 align:middle line:84% So couldn't be natural selection. 00:10:14.900 --> 00:10:20.420 align:middle line:84% And what's the third population genetic mechanism that could-- 00:10:20.420 --> 00:10:21.920 align:middle line:84% AUDIENCE: [INAUDIBLE] the third one. 00:10:21.920 --> 00:10:26.650 align:middle line:90% 00:10:26.650 --> 00:10:28.840 align:middle line:90% SCOTT WEISS: Genetic drift. 00:10:28.840 --> 00:10:32.830 align:middle line:84% So you had some mutant asthmatic that 00:10:32.830 --> 00:10:35.260 align:middle line:84% came into the American population, 00:10:35.260 --> 00:10:37.925 align:middle line:84% and over 20 years of time, they intermarried 00:10:37.925 --> 00:10:40.300 align:middle line:84% with all these other people and doubled the asthma rate-- 00:10:40.300 --> 00:10:42.113 align:middle line:90% plausible or implausible? 00:10:42.113 --> 00:10:43.030 align:middle line:90% AUDIENCE: Implausible. 00:10:43.030 --> 00:10:44.770 align:middle line:84% SCOTT WEISS: No, it can't happen. 00:10:44.770 --> 00:10:47.740 align:middle line:84% So you're right, it can't be genetic. 00:10:47.740 --> 00:10:50.265 align:middle line:90% So I give you half credit. 00:10:50.265 --> 00:10:54.110 align:middle line:90% 00:10:54.110 --> 00:10:56.660 align:middle line:84% I mean, most geneticists don't think like this, 00:10:56.660 --> 00:10:58.220 align:middle line:90% but they should think like this. 00:10:58.220 --> 00:11:02.060 align:middle line:84% The reality is that all of these genes 00:11:02.060 --> 00:11:05.150 align:middle line:84% operate in a developmental and an environmental context. 00:11:05.150 --> 00:11:06.500 align:middle line:90% All of your genes do. 00:11:06.500 --> 00:11:10.160 align:middle line:84% So the true underlying model for disease causation 00:11:10.160 --> 00:11:11.820 align:middle line:84% is gene by environment interaction. 00:11:11.820 --> 00:11:15.470 align:middle line:84% So it could very well be that there was some dramatic change 00:11:15.470 --> 00:11:18.050 align:middle line:84% in the environment, and now that's interacting with some 00:11:18.050 --> 00:11:20.870 align:middle line:84% other genes that it wasn't-- they weren't interacting before 00:11:20.870 --> 00:11:24.140 align:middle line:84% with, and now you've got this marked explosion in a number 00:11:24.140 --> 00:11:24.740 align:middle line:90% of cases. 00:11:24.740 --> 00:11:30.830 align:middle line:84% And that's almost-- certainly is the most comprehensive 00:11:30.830 --> 00:11:33.830 align:middle line:84% explanation, but it would have to devolve 00:11:33.830 --> 00:11:35.900 align:middle line:84% from some of environmental change, 00:11:35.900 --> 00:11:39.500 align:middle line:84% rather than some of primary change in the genes. 00:11:39.500 --> 00:11:42.050 align:middle line:84% But it could easily be that there is interaction 00:11:42.050 --> 00:11:44.330 align:middle line:84% between whatever the environmental exposure is 00:11:44.330 --> 00:11:47.420 align:middle line:84% and some of the underlying polymorphisms that 00:11:47.420 --> 00:11:49.670 align:middle line:84% may be disease related that are different now 00:11:49.670 --> 00:11:53.510 align:middle line:84% than they were back here when the disease rate was 00:11:53.510 --> 00:11:55.620 align:middle line:90% a lot lower. 00:11:55.620 --> 00:11:57.140 align:middle line:90% So big health problem. 00:11:57.140 --> 00:12:00.320 align:middle line:84% I don't want to dwell on this, because it's not 00:12:00.320 --> 00:12:05.120 align:middle line:84% the purpose of this course, but all this means 00:12:05.120 --> 00:12:07.730 align:middle line:84% is that people will give you money to study this, 00:12:07.730 --> 00:12:12.440 align:middle line:84% and they weren't so keen on doing that 20 or 30 years ago. 00:12:12.440 --> 00:12:14.180 align:middle line:84% The other important point is this disease 00:12:14.180 --> 00:12:16.140 align:middle line:90% is a disease of children. 00:12:16.140 --> 00:12:18.020 align:middle line:90% So you're going to think about-- 00:12:18.020 --> 00:12:21.260 align:middle line:84% I usually tell people, this is data 00:12:21.260 --> 00:12:24.410 align:middle line:84% looking at the age of onset in a closed 00:12:24.410 --> 00:12:26.600 align:middle line:90% population in Olmsted County. 00:12:26.600 --> 00:12:31.340 align:middle line:84% What's the famous medical center in Olmsted County, Minnesota? 00:12:31.340 --> 00:12:32.540 align:middle line:90% AUDIENCE: Mayo? 00:12:32.540 --> 00:12:33.710 align:middle line:84% SCOTT WEISS: You weren't supposed to answer that. 00:12:33.710 --> 00:12:35.340 align:middle line:90% They're supposed to answer that. 00:12:35.340 --> 00:12:37.200 align:middle line:90% AUDIENCE: [INAUDIBLE]. 00:12:37.200 --> 00:12:38.960 align:middle line:90% SCOTT WEISS: Mayo Clinic. 00:12:38.960 --> 00:12:41.270 align:middle line:90% So Mayo Clinic is-- 00:12:41.270 --> 00:12:44.600 align:middle line:84% everybody in Olmsted County goes to the Mayo Clinic. 00:12:44.600 --> 00:12:50.300 align:middle line:84% Now Saudi princes and sheikhs and famous people 00:12:50.300 --> 00:12:56.510 align:middle line:84% from all over the world and who's the King of Jordan-- 00:12:56.510 --> 00:13:00.350 align:middle line:90% King Hussein? 00:13:00.350 --> 00:13:02.930 align:middle line:90% He went there for his-- 00:13:02.930 --> 00:13:05.390 align:middle line:84% so Zach's mother, another famous person, 00:13:05.390 --> 00:13:06.802 align:middle line:90% went to the Mayo Clinic. 00:13:06.802 --> 00:13:08.510 align:middle line:84% So they get a lot of people from outside, 00:13:08.510 --> 00:13:10.580 align:middle line:84% but this data is based on the people who 00:13:10.580 --> 00:13:11.690 align:middle line:90% live in Olmsted County. 00:13:11.690 --> 00:13:14.420 align:middle line:84% Now if you live in Olmsted County, you don't go anywhere. 00:13:14.420 --> 00:13:16.440 align:middle line:84% You just sit right there and you stay there. 00:13:16.440 --> 00:13:19.410 align:middle line:84% So this is a fairly stable population. 00:13:19.410 --> 00:13:23.000 align:middle line:84% And they were able to capture all of the incident asthma 00:13:23.000 --> 00:13:28.190 align:middle line:84% cases and document them because they were all 00:13:28.190 --> 00:13:30.020 align:middle line:84% going to the Mayo Medical Center, 00:13:30.020 --> 00:13:31.730 align:middle line:84% and they had their chart records. 00:13:31.730 --> 00:13:35.240 align:middle line:84% So 90% of all of the people who were diagnosed 00:13:35.240 --> 00:13:37.940 align:middle line:84% as asthmatic in Olmsted County were diagnosed 00:13:37.940 --> 00:13:39.673 align:middle line:90% before the age of six. 00:13:39.673 --> 00:13:41.340 align:middle line:84% So this is a very, very important point. 00:13:41.340 --> 00:13:47.660 align:middle line:84% So this is the opposite of Alzheimer's disease 00:13:47.660 --> 00:13:50.030 align:middle line:84% because if you think about genetics, 00:13:50.030 --> 00:13:52.970 align:middle line:84% this is great for geneticists because I only 00:13:52.970 --> 00:13:56.150 align:middle line:84% have to wait six years from the time the kid is born, 00:13:56.150 --> 00:13:58.610 align:middle line:84% and I'm going to know whether they've got the disease 00:13:58.610 --> 00:13:59.420 align:middle line:90% phenotype or not. 00:13:59.420 --> 00:14:02.070 align:middle line:84% If I was waiting for Alzheimer's cases, 00:14:02.070 --> 00:14:04.010 align:middle line:90% it would be Waiting for Godot. 00:14:04.010 --> 00:14:09.380 align:middle line:84% I'd be waiting a long time before I'd get my cases. 00:14:09.380 --> 00:14:12.440 align:middle line:84% Now there's ways around that for the old people right. 00:14:12.440 --> 00:14:13.860 align:middle line:90% And what did geneticists do? 00:14:13.860 --> 00:14:16.480 align:middle line:84% How did they find the BRCA1 gene? 00:14:16.480 --> 00:14:20.630 align:middle line:84% What did they do to enhance the probability that you would 00:14:20.630 --> 00:14:24.110 align:middle line:84% find-- if you're looking at older people, 00:14:24.110 --> 00:14:28.940 align:middle line:84% how do you enrich for a genetic cause of a disease? 00:14:28.940 --> 00:14:30.668 align:middle line:90% What do you do? 00:14:30.668 --> 00:14:33.990 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:14:33.990 --> 00:14:37.202 align:middle line:90% 00:14:37.202 --> 00:14:39.660 align:middle line:84% SCOTT WEISS: Education doesn't have anything to do with it, 00:14:39.660 --> 00:14:42.420 align:middle line:90% I'm afraid. 00:14:42.420 --> 00:14:44.640 align:middle line:84% What characteristic of the cases would 00:14:44.640 --> 00:14:47.375 align:middle line:84% make you think it's more likely to be genetic? 00:14:47.375 --> 00:14:49.100 align:middle line:90% AUDIENCE: Family history. 00:14:49.100 --> 00:14:49.850 align:middle line:90% SCOTT WEISS: What? 00:14:49.850 --> 00:14:51.833 align:middle line:90% AUDIENCE: Family history. 00:14:51.833 --> 00:14:54.000 align:middle line:84% SCOTT WEISS: Family history, but what specifically-- 00:14:54.000 --> 00:14:55.150 align:middle line:90% the family history-- 00:14:55.150 --> 00:14:56.770 align:middle line:90% age of onset. 00:14:56.770 --> 00:14:59.060 align:middle line:84% You're looking for genetic causes of heart attacks, 00:14:59.060 --> 00:15:01.560 align:middle line:84% you're going to take the people that have heart attacks when 00:15:01.560 --> 00:15:02.220 align:middle line:90% they're age 50. 00:15:02.220 --> 00:15:05.670 align:middle line:84% So Ed Silverman, who's the world leader in COPD genetics, 00:15:05.670 --> 00:15:10.210 align:middle line:84% in my laboratory, is looking for early-onset COPD cases. 00:15:10.210 --> 00:15:13.860 align:middle line:84% So he gets cases where the age of onset 00:15:13.860 --> 00:15:17.860 align:middle line:90% is younger than the age of 52. 00:15:17.860 --> 00:15:19.258 align:middle line:90% So that's young. 00:15:19.258 --> 00:15:21.550 align:middle line:84% So if you were looking at Alzheimer's cases, you'd say, 00:15:21.550 --> 00:15:24.000 align:middle line:84% well, we want all of the cases of Alzheimer's people 00:15:24.000 --> 00:15:24.930 align:middle line:90% before the age of 60. 00:15:24.930 --> 00:15:27.600 align:middle line:84% And this is how Mary-Claire King found BRCA1. 00:15:27.600 --> 00:15:30.210 align:middle line:84% She looked at all of the early-onset breast cancer 00:15:30.210 --> 00:15:33.720 align:middle line:84% cases, people who got breast cancer in their 20s, their 30s, 00:15:33.720 --> 00:15:37.470 align:middle line:84% their 40s, instead of looking at older postmenopausal 00:15:37.470 --> 00:15:40.960 align:middle line:84% women, which is almost certainly another disease. 00:15:40.960 --> 00:15:42.780 align:middle line:84% So if you're looking at old people, 00:15:42.780 --> 00:15:44.730 align:middle line:84% one of the clever ways that geneticists 00:15:44.730 --> 00:15:50.370 align:middle line:84% enrich for genetic susceptibility 00:15:50.370 --> 00:15:53.050 align:middle line:84% is by looking at early age of onset. 00:15:53.050 --> 00:15:57.210 align:middle line:84% AUDIENCE: So they're selecting, potentially, 00:15:57.210 --> 00:15:59.310 align:middle line:84% a special case in a particular disease? 00:15:59.310 --> 00:16:00.850 align:middle line:90% SCOTT WEISS: Absolutely right. 00:16:00.850 --> 00:16:04.173 align:middle line:84% It's a little bit like, in some perverse kind of way, 00:16:04.173 --> 00:16:06.090 align:middle line:84% it's a little bit like the geniculate isolate. 00:16:06.090 --> 00:16:11.580 align:middle line:84% You might find a gene that is specific for that particular 00:16:11.580 --> 00:16:15.100 align:middle line:90% type of early onset disease. 00:16:15.100 --> 00:16:18.330 align:middle line:84% So you find a gene for early onset Alzheimer's, but not 00:16:18.330 --> 00:16:21.810 align:middle line:84% for garden variety old age Alzheimer's 00:16:21.810 --> 00:16:26.850 align:middle line:84% that occurs in virtually everybody by the time 00:16:26.850 --> 00:16:28.590 align:middle line:90% they're 90. 00:16:28.590 --> 00:16:30.900 align:middle line:90% So you're right. 00:16:30.900 --> 00:16:36.030 align:middle line:84% But we're still in, I think, what geneticists would 00:16:36.030 --> 00:16:38.010 align:middle line:90% say the early stages of this. 00:16:38.010 --> 00:16:41.100 align:middle line:84% And because we're still in the early stages, most of us 00:16:41.100 --> 00:16:43.950 align:middle line:84% would be happy if we found any gene. 00:16:43.950 --> 00:16:47.010 align:middle line:84% So you're going to be in science, 00:16:47.010 --> 00:16:48.913 align:middle line:84% if you find that early-onset gene, 00:16:48.913 --> 00:16:50.580 align:middle line:84% and nobody's going to be criticizing you 00:16:50.580 --> 00:16:54.720 align:middle line:84% because it's not the gene for all breast cancer or all 00:16:54.720 --> 00:16:55.530 align:middle line:90% Alzheimer's. 00:16:55.530 --> 00:16:57.270 align:middle line:84% AUDIENCE: [? So yet ?] another example of low-hanging fruit. 00:16:57.270 --> 00:16:58.145 align:middle line:90% SCOTT WEISS: Exactly. 00:16:58.145 --> 00:17:00.330 align:middle line:90% 00:17:00.330 --> 00:17:02.670 align:middle line:84% So a little bit more about the disease-- 00:17:02.670 --> 00:17:04.770 align:middle line:90% most of the kids are allergic. 00:17:04.770 --> 00:17:08.670 align:middle line:84% Allergy is probably the big reason why the asthma epidemic 00:17:08.670 --> 00:17:15.390 align:middle line:84% occurred, and that means that they have this particular type 00:17:15.390 --> 00:17:18.630 align:middle line:84% of an inflammatory process where antigen presented 00:17:18.630 --> 00:17:21.119 align:middle line:84% to dendritic cells in the airways 00:17:21.119 --> 00:17:25.349 align:middle line:84% activate these CD4 positive T lymphocytes, which then 00:17:25.349 --> 00:17:29.070 align:middle line:84% elaborate this series of inflammatory cytokines, which 00:17:29.070 --> 00:17:32.520 align:middle line:84% go to these inflammatory cells, which then infiltrate 00:17:32.520 --> 00:17:34.950 align:middle line:84% the airways and set up an inflammatory reaction 00:17:34.950 --> 00:17:39.690 align:middle line:84% with coughing, wheezing, airways responsiveness, et cetera. 00:17:39.690 --> 00:17:41.550 align:middle line:84% This is all well-known, but it does 00:17:41.550 --> 00:17:44.940 align:middle line:84% suggest a whole host of other potential phenotypes 00:17:44.940 --> 00:17:47.100 align:middle line:84% that you could potentially look at. 00:17:47.100 --> 00:17:50.460 align:middle line:84% And it also gets at this concept of ontogeny 00:17:50.460 --> 00:17:56.190 align:middle line:84% of the immune system, where T null cells, at some point, 00:17:56.190 --> 00:17:58.990 align:middle line:84% differentiate into these Th1 and Th2 cells, 00:17:58.990 --> 00:18:01.740 align:middle line:84% which are determined-- their phenotype 00:18:01.740 --> 00:18:05.280 align:middle line:84% is determined by which cytokines they actually elaborate. 00:18:05.280 --> 00:18:06.870 align:middle line:84% And I've got a question mark here, 00:18:06.870 --> 00:18:11.280 align:middle line:84% but, actually, this particular step, which is the crosstalk 00:18:11.280 --> 00:18:14.550 align:middle line:84% and interaction between these two types of cells that 00:18:14.550 --> 00:18:18.900 align:middle line:84% are controlled by two specific genes that elaborate 00:18:18.900 --> 00:18:19.680 align:middle line:90% cytokines-- 00:18:19.680 --> 00:18:27.990 align:middle line:84% IL-10 and TGF beta, and we genotyped both of those genes 00:18:27.990 --> 00:18:34.380 align:middle line:84% in asthma and COPD, and they're important in both diseases. 00:18:34.380 --> 00:18:36.300 align:middle line:84% Now it's important for you to understand 00:18:36.300 --> 00:18:38.280 align:middle line:84% that I skewed things a little bit 00:18:38.280 --> 00:18:42.150 align:middle line:84% because I told you that asthma is a Th2 disease, 00:18:42.150 --> 00:18:44.850 align:middle line:90% and Th2 diseases have increased. 00:18:44.850 --> 00:18:47.130 align:middle line:84% There's this increase in allergic rhinitis, food 00:18:47.130 --> 00:18:49.308 align:middle line:90% allergy, asthma, et cetera. 00:18:49.308 --> 00:18:51.600 align:middle line:84% This novel gene that I was telling you about just a few 00:18:51.600 --> 00:18:54.330 align:middle line:84% minutes ago-- it's going to come out in Science on Friday-- 00:18:54.330 --> 00:18:59.430 align:middle line:84% that gene is expressed in the skin and in gut epithelium 00:18:59.430 --> 00:19:01.320 align:middle line:84% and in airway epithelium, suggesting 00:19:01.320 --> 00:19:03.960 align:middle line:84% that may be important in all these different types 00:19:03.960 --> 00:19:06.360 align:middle line:84% of allergic diseases which, again, has heightened 00:19:06.360 --> 00:19:12.150 align:middle line:84% people's interest in the gene and its potential importance. 00:19:12.150 --> 00:19:18.030 align:middle line:84% But it's also important to recognize that Th1 diseases 00:19:18.030 --> 00:19:19.030 align:middle line:90% have also increased. 00:19:19.030 --> 00:19:22.290 align:middle line:84% So give me some examples of some Th1 diseases. 00:19:22.290 --> 00:19:23.790 align:middle line:84% So the epidemiology-- and the reason 00:19:23.790 --> 00:19:26.160 align:middle line:84% I'm bringing this up is that most 00:19:26.160 --> 00:19:28.200 align:middle line:84% of the immunology community is focused on-- this 00:19:28.200 --> 00:19:30.480 align:middle line:84% is why it's important if you're going to be a good geneticist, 00:19:30.480 --> 00:19:32.190 align:middle line:84% you've got to really know your disease. 00:19:32.190 --> 00:19:35.100 align:middle line:84% You can't just wave your hand at it and say-- 00:19:35.100 --> 00:19:40.520 align:middle line:84% and I think that age of the generalist geneticist, 00:19:40.520 --> 00:19:42.290 align:middle line:84% geneticists that sort of, oh, you know, 00:19:42.290 --> 00:19:46.940 align:middle line:84% I'm going to study this disease and I'm to study that disease-- 00:19:46.940 --> 00:19:47.948 align:middle line:90% not with complex traits. 00:19:47.948 --> 00:19:48.990 align:middle line:90% That's not going to work. 00:19:48.990 --> 00:19:50.630 align:middle line:84% You're going to have to really know your disease because you 00:19:50.630 --> 00:19:51.590 align:middle line:84% have to know the environment, you're 00:19:51.590 --> 00:19:53.630 align:middle line:84% going to have to know the natural history, 00:19:53.630 --> 00:19:56.005 align:middle line:84% you're going to have to know the intermediate phenotypes, 00:19:56.005 --> 00:20:01.070 align:middle line:84% and you have to really understand the biology as well. 00:20:01.070 --> 00:20:03.200 align:middle line:84% The point here is these Th1 diseases-- give me 00:20:03.200 --> 00:20:05.360 align:middle line:90% an example of a Th1 disease. 00:20:05.360 --> 00:20:07.950 align:middle line:84% AUDIENCE: Well, I'm thinking about that-- 00:20:07.950 --> 00:20:10.940 align:middle line:84% could you tell me the autoimmune diseases, 00:20:10.940 --> 00:20:13.730 align:middle line:84% like inflammatory bowel disease, what are they [INAUDIBLE]?? 00:20:13.730 --> 00:20:14.772 align:middle line:90% SCOTT WEISS: They're Th1. 00:20:14.772 --> 00:20:16.550 align:middle line:90% So Crohn's disease-- 00:20:16.550 --> 00:20:17.930 align:middle line:90% Th1 disease. 00:20:17.930 --> 00:20:20.360 align:middle line:84% Juvenile rheumatoid arthritis-- Th1 disease. 00:20:20.360 --> 00:20:25.370 align:middle line:90% 00:20:25.370 --> 00:20:27.440 align:middle line:90% Psoriasis-- Th1 disease. 00:20:27.440 --> 00:20:30.860 align:middle line:90% 00:20:30.860 --> 00:20:35.520 align:middle line:90% Juvenile diabetes-- Th1 disease. 00:20:35.520 --> 00:20:39.470 align:middle line:84% And the reason-- so if the prevalence of these 00:20:39.470 --> 00:20:42.470 align:middle line:84% has gone up, and the prevalence of these has gone up, 00:20:42.470 --> 00:20:46.400 align:middle line:84% people are thinking that there's something going on further 00:20:46.400 --> 00:20:51.680 align:middle line:84% up here that has to do with Treg cells, 00:20:51.680 --> 00:20:55.220 align:middle line:84% cells that regulate T cells in terms of their differentiation, 00:20:55.220 --> 00:20:58.340 align:middle line:84% because it can't be just at this level 00:20:58.340 --> 00:21:00.230 align:middle line:90% that the immunologic defect is. 00:21:00.230 --> 00:21:04.580 align:middle line:84% So it raises the possibility that there are genes-- 00:21:04.580 --> 00:21:11.270 align:middle line:84% a Foxp3 getter, T-bet, a whole bunch of other genes 00:21:11.270 --> 00:21:15.170 align:middle line:84% that are proximal to the Th1, Th2, C4 00:21:15.170 --> 00:21:17.660 align:middle line:84% lymphocytes that may be important in all 00:21:17.660 --> 00:21:21.560 align:middle line:90% of these autoimmune diseases. 00:21:21.560 --> 00:21:25.460 align:middle line:84% And people are just now starting to look at that. 00:21:25.460 --> 00:21:28.040 align:middle line:84% And, obviously, the environmental and genetic 00:21:28.040 --> 00:21:31.040 align:middle line:84% factors that influence the differentiation 00:21:31.040 --> 00:21:37.640 align:middle line:84% of the immune system, or how do people actually 00:21:37.640 --> 00:21:40.790 align:middle line:90% tolerize the foreign antigen-- 00:21:40.790 --> 00:21:46.050 align:middle line:84% that's the kind of really simple complicated question 00:21:46.050 --> 00:21:48.050 align:middle line:84% that, if you could figure out an answer to that, 00:21:48.050 --> 00:21:50.450 align:middle line:90% you'd win a Nobel Prize. 00:21:50.450 --> 00:21:54.180 align:middle line:84% So that's what my laboratory is starting to work on. 00:21:54.180 --> 00:21:57.817 align:middle line:84% So this is just to show you, again, 00:21:57.817 --> 00:21:59.400 align:middle line:84% what I've already told you, that there 00:21:59.400 --> 00:22:04.580 align:middle line:84% are a bunch of factors, mostly bacteria and viruses 00:22:04.580 --> 00:22:10.020 align:middle line:84% and parasites that influence this Th1/Th2 differentiation. 00:22:10.020 --> 00:22:14.270 align:middle line:84% And environmental factors that influence those things 00:22:14.270 --> 00:22:15.800 align:middle line:90% are presumed to be important. 00:22:15.800 --> 00:22:18.290 align:middle line:84% And one would want to know both the genes 00:22:18.290 --> 00:22:20.480 align:middle line:84% and the environmental factors that are involved 00:22:20.480 --> 00:22:22.280 align:middle line:90% in this particular disease. 00:22:22.280 --> 00:22:26.480 align:middle line:84% There happened to be a whole host of environmental factors 00:22:26.480 --> 00:22:30.540 align:middle line:84% that are correlates of those sorts of changes. 00:22:30.540 --> 00:22:33.872 align:middle line:84% And I've listed a bunch of them here. 00:22:33.872 --> 00:22:35.330 align:middle line:84% We get very interested in this one. 00:22:35.330 --> 00:22:39.530 align:middle line:84% We went to China in 1996 to do an asthma genetics study, 00:22:39.530 --> 00:22:42.950 align:middle line:84% and I noticed how different the environment was there. 00:22:42.950 --> 00:22:47.900 align:middle line:84% And this left-hand category would 00:22:47.900 --> 00:22:50.450 align:middle line:84% summarize what you would see if you were standing 00:22:50.450 --> 00:22:53.270 align:middle line:84% in rural China and what you would see in terms 00:22:53.270 --> 00:22:56.900 align:middle line:90% of the environmental exposures. 00:22:56.900 --> 00:23:01.690 align:middle line:84% Very, very low asthma rates in rural China-- it's about 1%. 00:23:01.690 --> 00:23:05.270 align:middle line:84% And it is a progressive increase in gradient in terms of disease 00:23:05.270 --> 00:23:08.970 align:middle line:84% prevalence as you march towards Beijing or Shanghai-- much, 00:23:08.970 --> 00:23:10.790 align:middle line:90% much higher rates. 00:23:10.790 --> 00:23:13.400 align:middle line:84% AUDIENCE: Let me ask you some questions. 00:23:13.400 --> 00:23:17.302 align:middle line:84% Is it not factually wrong that Chinese populations-- 00:23:17.302 --> 00:23:19.010 align:middle line:84% I thought the families were [INAUDIBLE]?? 00:23:19.010 --> 00:23:19.880 align:middle line:90% SCOTT WEISS: No. 00:23:19.880 --> 00:23:24.260 align:middle line:84% See, that's again-- a little bit of knowledge is a bad thing. 00:23:24.260 --> 00:23:27.290 align:middle line:84% If you get into rural China, actually, 00:23:27.290 --> 00:23:33.308 align:middle line:84% where farming is what everybody does-- 00:23:33.308 --> 00:23:34.850 align:middle line:84% although the central government would 00:23:34.850 --> 00:23:40.970 align:middle line:84% say there's a two-child policy or a one-child policy, 00:23:40.970 --> 00:23:44.480 align:middle line:84% in rural China, they just they have as many children as they-- 00:23:44.480 --> 00:23:47.000 align:middle line:84% they may not register them with Social Security, 00:23:47.000 --> 00:23:49.640 align:middle line:84% but if they need three kids to run the farm or four kids, 00:23:49.640 --> 00:23:51.750 align:middle line:84% they have as many kids as they want. 00:23:51.750 --> 00:23:54.320 align:middle line:84% So we found a lot of families with four or five, 00:23:54.320 --> 00:23:55.940 align:middle line:90% six, eight kids. 00:23:55.940 --> 00:23:58.863 align:middle line:84% AUDIENCE: All right, here we go, [INAUDIBLE].. 00:23:58.863 --> 00:24:00.530 align:middle line:84% SCOTT WEISS: So why has it been presumed 00:24:00.530 --> 00:24:02.780 align:middle line:84% to be difficult to do this kind of work? 00:24:02.780 --> 00:24:05.060 align:middle line:84% What's the reason that it's difficult? 00:24:05.060 --> 00:24:07.430 align:middle line:84% And I think these are some of the reasons. 00:24:07.430 --> 00:24:09.980 align:middle line:84% And some of them relate to the issues of study design-- 00:24:09.980 --> 00:24:11.630 align:middle line:84% the things that we were talking about. 00:24:11.630 --> 00:24:14.780 align:middle line:84% One is this whole idea of genetic heterogeneity. 00:24:14.780 --> 00:24:17.030 align:middle line:84% Particularly, if the underlying model here 00:24:17.030 --> 00:24:19.130 align:middle line:84% is gene by environment interaction, 00:24:19.130 --> 00:24:22.590 align:middle line:84% presumably you could get the same phenotype-- 00:24:22.590 --> 00:24:26.440 align:middle line:84% and these phenotypes are determined by multiple genes. 00:24:26.440 --> 00:24:29.480 align:middle line:84% You could get the same phenotype, either high IgE 00:24:29.480 --> 00:24:32.570 align:middle line:84% or airways responsiveness in population A 00:24:32.570 --> 00:24:35.780 align:middle line:84% with a very different constellation of genes 00:24:35.780 --> 00:24:37.670 align:middle line:84% and environmental exposures, and you 00:24:37.670 --> 00:24:40.760 align:middle line:84% can get the same thing in population B 00:24:40.760 --> 00:24:44.240 align:middle line:90% with different genes. 00:24:44.240 --> 00:24:47.180 align:middle line:84% So this is the geniculate heterogeneity thing 00:24:47.180 --> 00:24:50.270 align:middle line:84% is a reason for focusing on a genetic isolate. 00:24:50.270 --> 00:24:53.100 align:middle line:84% But then you have to worry about the generalizability question. 00:24:53.100 --> 00:24:56.188 align:middle line:84% So, in fact, in asthma there are four positionally cloned genes, 00:24:56.188 --> 00:24:57.980 align:middle line:84% counting the paper that's going to come out 00:24:57.980 --> 00:25:01.490 align:middle line:84% on Friday in Science, and of those four, 00:25:01.490 --> 00:25:04.040 align:middle line:84% the first is the only one that's people 00:25:04.040 --> 00:25:06.620 align:middle line:84% have really attempted to replicate, 00:25:06.620 --> 00:25:08.188 align:middle line:90% and it's gotten mixed results. 00:25:08.188 --> 00:25:10.730 align:middle line:84% There are some people that have replicated it and some people 00:25:10.730 --> 00:25:11.430 align:middle line:90% that haven't. 00:25:11.430 --> 00:25:13.610 align:middle line:84% So it's one of those genes that probably 00:25:13.610 --> 00:25:16.940 align:middle line:84% falls into this category of, well, it's not a major gene. 00:25:16.940 --> 00:25:18.260 align:middle line:90% It's a minor gene. 00:25:18.260 --> 00:25:21.380 align:middle line:84% It's one of the 200 genes that determine asthma, 00:25:21.380 --> 00:25:24.710 align:middle line:84% but it's not one of the top 10 in every population. 00:25:24.710 --> 00:25:28.010 align:middle line:84% AUDIENCE: What's your guess about this new gene 00:25:28.010 --> 00:25:29.377 align:middle line:90% is going to be [INAUDIBLE]. 00:25:29.377 --> 00:25:31.460 align:middle line:84% SCOTT WEISS: My guess my guess about this new gene 00:25:31.460 --> 00:25:34.940 align:middle line:90% is that it's a major player. 00:25:34.940 --> 00:25:43.370 align:middle line:84% But having said that, the point that I made to the Science 00:25:43.370 --> 00:25:50.690 align:middle line:84% writer who was doing this is it that's what science is all 00:25:50.690 --> 00:25:54.890 align:middle line:84% about is replicating this, seeing how important it really 00:25:54.890 --> 00:25:56.540 align:middle line:84% is, and seeing what actually happens. 00:25:56.540 --> 00:26:03.350 align:middle line:84% I mean I think you can get a clue as to whether you've 00:26:03.350 --> 00:26:06.110 align:middle line:84% got hold of an area where there's a potential major locus 00:26:06.110 --> 00:26:08.750 align:middle line:84% or not, by looking at the replicability 00:26:08.750 --> 00:26:11.510 align:middle line:84% of the linkage peaks in a particular region 00:26:11.510 --> 00:26:13.110 align:middle line:84% for a complex trait-- in other words, 00:26:13.110 --> 00:26:15.740 align:middle line:84% if you've got a region where there's a linkage peak, 00:26:15.740 --> 00:26:18.020 align:middle line:84% and there are 10 different studies and 10 00:26:18.020 --> 00:26:20.180 align:middle line:84% different populations, and there's always 00:26:20.180 --> 00:26:24.020 align:middle line:84% a peak in the same region, then the chances 00:26:24.020 --> 00:26:26.090 align:middle line:84% are that there's a major gene in there that's 00:26:26.090 --> 00:26:28.760 align:middle line:84% probably going to apply to a bunch of different populations. 00:26:28.760 --> 00:26:31.700 align:middle line:84% Well, this is-- going back to the Science article-- 00:26:31.700 --> 00:26:34.640 align:middle line:84% this is a region where there have been a number of people 00:26:34.640 --> 00:26:38.640 align:middle line:90% have found a peak there. 00:26:38.640 --> 00:26:43.160 align:middle line:84% The other problem here is that unlike single gene disorders, 00:26:43.160 --> 00:26:45.260 align:middle line:84% where there's a known mode of inheritance, 00:26:45.260 --> 00:26:48.380 align:middle line:84% you get everything under the kitchen sink here. 00:26:48.380 --> 00:26:51.290 align:middle line:84% So you get some of these genes that are autosomal recessive, 00:26:51.290 --> 00:26:55.950 align:middle line:84% and some are autosomal dominance, and some are-- 00:26:55.950 --> 00:26:59.810 align:middle line:84% so you're getting a whole bunch of things jumbled up 00:26:59.810 --> 00:27:02.820 align:middle line:84% in one phenotype, which makes it very difficult. 00:27:02.820 --> 00:27:05.840 align:middle line:84% And then there's this problem of phenocopies. 00:27:05.840 --> 00:27:07.790 align:middle line:90% So what's a phenocopy? 00:27:07.790 --> 00:27:09.620 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:27:09.620 --> 00:27:10.250 align:middle line:90% 00:27:10.250 --> 00:27:16.310 align:middle line:84% SCOTT WEISS: Give me an example from your own clinical 00:27:16.310 --> 00:27:18.460 align:middle line:90% experience of somebody who's-- 00:27:18.460 --> 00:27:21.380 align:middle line:90% a phenocopy who's not due to a-- 00:27:21.380 --> 00:27:25.210 align:middle line:90% 00:27:25.210 --> 00:27:27.355 align:middle line:84% well, just like can get these diseases from genes, 00:27:27.355 --> 00:27:29.480 align:middle line:84% you can get them from exposures in the environment. 00:27:29.480 --> 00:27:32.620 align:middle line:84% So what if you've got some guy who smokes four packs a day 00:27:32.620 --> 00:27:36.190 align:middle line:84% and he's 50 years old and he has a whopping big heart attack? 00:27:36.190 --> 00:27:39.670 align:middle line:84% Well, maybe when you quiz him, and he has no family history, 00:27:39.670 --> 00:27:41.110 align:middle line:90% but you smoke four packs a day. 00:27:41.110 --> 00:27:44.380 align:middle line:84% Well, you can get a heart attack from smoking four packs a day, 00:27:44.380 --> 00:27:47.110 align:middle line:84% and you don't need to have any genes at all for heart attack. 00:27:47.110 --> 00:27:49.060 align:middle line:84% You can just-- so that's a phenocopy. 00:27:49.060 --> 00:27:51.430 align:middle line:84% He's going to look like somebody who's 00:27:51.430 --> 00:27:56.290 align:middle line:84% a genetic susceptible because had a heart attack at 50 years 00:27:56.290 --> 00:28:01.020 align:middle line:84% old, but it's all due to an environmental exposure. 00:28:01.020 --> 00:28:04.090 align:middle line:90% And complete penetrance. 00:28:04.090 --> 00:28:08.740 align:middle line:84% So this is a problem even in single gene disorders 00:28:08.740 --> 00:28:11.440 align:middle line:84% because they're clearly examples-- 00:28:11.440 --> 00:28:16.120 align:middle line:84% hemochromatosis, cystic fibrosis-- very different 00:28:16.120 --> 00:28:18.790 align:middle line:84% spectrum of diseases in these-- we 00:28:18.790 --> 00:28:26.410 align:middle line:84% know that the CFTR gene causes cystic fibrosis. 00:28:26.410 --> 00:28:28.660 align:middle line:84% You've got some people who have completely normal lung 00:28:28.660 --> 00:28:32.530 align:middle line:84% function, no lung disease at all, and all they've got 00:28:32.530 --> 00:28:34.570 align:middle line:84% is mild pancreatic insufficiency, 00:28:34.570 --> 00:28:36.610 align:middle line:84% and you've got other people who are 00:28:36.610 --> 00:28:38.720 align:middle line:90% totally debilitated from it. 00:28:38.720 --> 00:28:40.932 align:middle line:84% So part of that can be penetrance. 00:28:40.932 --> 00:28:42.640 align:middle line:84% Part of it can be environmental exposure. 00:28:42.640 --> 00:28:44.800 align:middle line:84% But incomplete penetrance is important. 00:28:44.800 --> 00:28:47.440 align:middle line:84% Then you've got this problem of multiple genes. 00:28:47.440 --> 00:28:53.890 align:middle line:84% People have very-- the lay public has a very delusional-- 00:28:53.890 --> 00:28:56.260 align:middle line:84% they think the genes are immutable. 00:28:56.260 --> 00:28:59.130 align:middle line:84% If you've got those genes, that's it. 00:28:59.130 --> 00:29:00.610 align:middle line:90% It can't be changed. 00:29:00.610 --> 00:29:02.350 align:middle line:84% And they also-- they're monolithic. 00:29:02.350 --> 00:29:05.680 align:middle line:84% They're really big, whereas the reality is that anyone-- 00:29:05.680 --> 00:29:08.320 align:middle line:84% you've got 33,000 genes in the genome, 00:29:08.320 --> 00:29:10.930 align:middle line:84% take a disease like asthma, which isn't very complicated-- 00:29:10.930 --> 00:29:13.390 align:middle line:90% maybe there are 200-250-- 00:29:13.390 --> 00:29:15.100 align:middle line:90% I don't know, a lot-- 00:29:15.100 --> 00:29:16.780 align:middle line:90% that are probably important. 00:29:16.780 --> 00:29:23.680 align:middle line:84% Maybe 10 to play a role in most every population. 00:29:23.680 --> 00:29:25.575 align:middle line:84% And a lot of environmental things going on, 00:29:25.575 --> 00:29:26.950 align:middle line:84% and it makes it very complicated, 00:29:26.950 --> 00:29:29.600 align:middle line:84% and that's why guys like this guy 00:29:29.600 --> 00:29:31.900 align:middle line:84% are going to make big bucks because they're 00:29:31.900 --> 00:29:37.000 align:middle line:84% going to be able to model all of the different pathways and all 00:29:37.000 --> 00:29:39.040 align:middle line:84% the different genes together in some more 00:29:39.040 --> 00:29:43.300 align:middle line:84% realistic model of systems biology 00:29:43.300 --> 00:29:45.970 align:middle line:84% or some actual way of looking at this. 00:29:45.970 --> 00:29:49.810 align:middle line:84% But the point here basically is look it's 00:29:49.810 --> 00:29:51.655 align:middle line:90% complicated to do this stuff. 00:29:51.655 --> 00:29:53.530 align:middle line:84% But again, going back to what I said earlier, 00:29:53.530 --> 00:29:54.680 align:middle line:90% it's getting a lot easier. 00:29:54.680 --> 00:30:00.400 align:middle line:84% So 2002-- one position we cloned gene for asthma. 00:30:00.400 --> 00:30:05.170 align:middle line:84% 2003-- two positionally cloned gene from asthma. 00:30:05.170 --> 00:30:08.290 align:middle line:84% 2004-- first paper is already out, 00:30:08.290 --> 00:30:11.220 align:middle line:84% and there's probably going to be four or five more. 00:30:11.220 --> 00:30:13.720 align:middle line:84% So it's going to be four or five this year, year after that, 00:30:13.720 --> 00:30:15.190 align:middle line:90% they're going to be probably 10. 00:30:15.190 --> 00:30:16.648 align:middle line:84% And all of a sudden now, you've got 00:30:16.648 --> 00:30:19.300 align:middle line:84% 20 genes identified for the disease by positional cloning. 00:30:19.300 --> 00:30:22.210 align:middle line:84% And that is the history of complex trait genetics, 00:30:22.210 --> 00:30:23.260 align:middle line:90% and it's going to be-- 00:30:23.260 --> 00:30:25.450 align:middle line:90% it's happening right now. 00:30:25.450 --> 00:30:28.750 align:middle line:84% Right this very moment, all across the world, 00:30:28.750 --> 00:30:31.930 align:middle line:84% labs like mine are right in the middle of the fray 00:30:31.930 --> 00:30:33.670 align:middle line:90% doing this stuff. 00:30:33.670 --> 00:30:38.020 align:middle line:84% This is, simply put, the single most exciting time 00:30:38.020 --> 00:30:39.160 align:middle line:90% to be doing human genetics. 00:30:39.160 --> 00:30:43.070 align:middle line:84% And it's going to go on for a while, 00:30:43.070 --> 00:30:46.370 align:middle line:90% but who knows for how long? 00:30:46.370 --> 00:30:48.070 align:middle line:84% So then there's this other problem 00:30:48.070 --> 00:30:51.040 align:middle line:84% of pleiotropy, which can have one gene 00:30:51.040 --> 00:30:53.020 align:middle line:84% and it can do a lot of different things. 00:30:53.020 --> 00:30:54.550 align:middle line:84% You got the CFTR gene that gives you 00:30:54.550 --> 00:30:57.610 align:middle line:84% lung disease, pancreatic insufficiency, and fertility. 00:30:57.610 --> 00:31:03.340 align:middle line:84% It all has to do with mucosa and epithelia and different organ 00:31:03.340 --> 00:31:05.830 align:middle line:84% systems where this particular genus expressed. 00:31:05.830 --> 00:31:09.040 align:middle line:84% So one of the genes we're looking at-- 00:31:09.040 --> 00:31:12.580 align:middle line:84% did Jeff talk to you about CRHR1? 00:31:12.580 --> 00:31:16.540 align:middle line:84% Did he show you the data about CRHR1 last week? 00:31:16.540 --> 00:31:22.810 align:middle line:84% So that gene is expressed-- is that gene 00:31:22.810 --> 00:31:23.860 align:middle line:90% expressed in the lung? 00:31:23.860 --> 00:31:26.930 align:middle line:90% 00:31:26.930 --> 00:31:29.965 align:middle line:90% Yes or no? 00:31:29.965 --> 00:31:31.800 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:31:31.800 --> 00:31:32.570 align:middle line:90% SCOTT WEISS: No. 00:31:32.570 --> 00:31:35.180 align:middle line:90% Not expressed in the lung. 00:31:35.180 --> 00:31:39.770 align:middle line:84% It's the receptor for CRF or CRH, 00:31:39.770 --> 00:31:42.330 align:middle line:90% and it's expressed in the brain. 00:31:42.330 --> 00:31:46.490 align:middle line:84% So what other disease might that gene potentially 00:31:46.490 --> 00:31:47.360 align:middle line:90% be important in? 00:31:47.360 --> 00:31:50.440 align:middle line:90% 00:31:50.440 --> 00:31:51.440 align:middle line:90% He's an endocrinologist. 00:31:51.440 --> 00:31:55.220 align:middle line:90% He's forbidden from answering. 00:31:55.220 --> 00:31:56.415 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:31:56.415 --> 00:31:57.040 align:middle line:90% AUDIENCE: What? 00:31:57.040 --> 00:31:58.910 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:31:58.910 --> 00:32:01.160 align:middle line:84% SCOTT WEISS: It comes from the hypothalamus, actually. 00:32:01.160 --> 00:32:07.130 align:middle line:90% 00:32:07.130 --> 00:32:08.905 align:middle line:90% AUDIENCE: What clinical disease? 00:32:08.905 --> 00:32:10.025 align:middle line:90% Well, I just-- 00:32:10.025 --> 00:32:11.120 align:middle line:90% [INAUDIBLE] hypertension. 00:32:11.120 --> 00:32:12.203 align:middle line:90% SCOTT WEISS: Hypertension? 00:32:12.203 --> 00:32:13.850 align:middle line:90% AUDIENCE: It's not-- 00:32:13.850 --> 00:32:17.150 align:middle line:84% SCOTT WEISS: There are endocrine causes of hypertension. 00:32:17.150 --> 00:32:19.460 align:middle line:84% AUDIENCE: [INAUDIBLE] course yet? 00:32:19.460 --> 00:32:22.490 align:middle line:84% SCOTT WEISS: So anybody had physiology? 00:32:22.490 --> 00:32:23.570 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:32:23.570 --> 00:32:24.620 align:middle line:90% SCOTT WEISS: HPA? 00:32:24.620 --> 00:32:28.342 align:middle line:90% AUDIENCE: OK, so [INAUDIBLE]? 00:32:28.342 --> 00:32:30.050 align:middle line:84% AUDIENCE: No, think common diseases, man. 00:32:30.050 --> 00:32:33.045 align:middle line:90% SCOTT WEISS: Common disease. 00:32:33.045 --> 00:32:33.920 align:middle line:90% [INAUDIBLE] tell you. 00:32:33.920 --> 00:32:37.010 align:middle line:90% It's depression. 00:32:37.010 --> 00:32:40.670 align:middle line:84% It's been studied a huge amount in the section 00:32:40.670 --> 00:32:44.445 align:middle line:84% brains of people who committed suicide, and this-- 00:32:44.445 --> 00:32:45.320 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:32:45.320 --> 00:32:46.987 align:middle line:84% SCOTT WEISS: I mean, all kinds of things 00:32:46.987 --> 00:32:52.490 align:middle line:84% show that CRF and CRHR1, which is the ligand and the receptor 00:32:52.490 --> 00:32:56.270 align:middle line:84% are important in affective disorders. 00:32:56.270 --> 00:32:58.680 align:middle line:84% AUDIENCE: Is there any link to [INAUDIBLE]?? 00:32:58.680 --> 00:33:02.090 align:middle line:84% SCOTT WEISS: Well, there's an association between 00:33:02.090 --> 00:33:04.730 align:middle line:84% our [INAUDIBLE] type and depression, 00:33:04.730 --> 00:33:08.372 align:middle line:84% and Julia [INAUDIBLE] Mexican-American-- 00:33:08.372 --> 00:33:09.080 align:middle line:90% AUDIENCE: Really? 00:33:09.080 --> 00:33:10.167 align:middle line:90% SCOTT WEISS: Yeah. 00:33:10.167 --> 00:33:11.000 align:middle line:90% AUDIENCE: Very cool. 00:33:11.000 --> 00:33:14.610 align:middle line:84% SCOTT WEISS: Yeah, so that's pleiotropy. 00:33:14.610 --> 00:33:17.120 align:middle line:84% And then, obviously, you've got this problem 00:33:17.120 --> 00:33:21.470 align:middle line:90% of penetrance, which is-- 00:33:21.470 --> 00:33:23.630 align:middle line:84% individuals with a genotype who actually 00:33:23.630 --> 00:33:28.760 align:middle line:84% express the trait, and IgE genes can be important in hay fever. 00:33:28.760 --> 00:33:30.930 align:middle line:90% They can be important in asthma. 00:33:30.930 --> 00:33:32.570 align:middle line:84% And there are some people who don't 00:33:32.570 --> 00:33:35.810 align:middle line:84% have high IgE at all, even though they've got the genes. 00:33:35.810 --> 00:33:39.020 align:middle line:90% 00:33:39.020 --> 00:33:41.330 align:middle line:84% And these are some other examples of things like-- 00:33:41.330 --> 00:33:43.250 align:middle line:84% The basic point I'm trying to make here 00:33:43.250 --> 00:33:46.460 align:middle line:84% is that these are reasons that have been given for why 00:33:46.460 --> 00:33:48.233 align:middle line:90% doing this stuff is hard. 00:33:48.233 --> 00:33:49.400 align:middle line:90% But I'll tell you something. 00:33:49.400 --> 00:33:52.070 align:middle line:84% Really, the hard part has been developing 00:33:52.070 --> 00:33:54.950 align:middle line:84% the bioinformatics infrastructure, the tools, 00:33:54.950 --> 00:34:00.410 align:middle line:84% the bioinformatics tools, and cheap, reliable genotyping. 00:34:00.410 --> 00:34:03.710 align:middle line:84% Those have really been things that have been important. 00:34:03.710 --> 00:34:07.430 align:middle line:84% And just in the little bit of time that I've been doing this, 00:34:07.430 --> 00:34:13.880 align:middle line:84% my genotyping costs have gone from $1.20 a SNP genotype, 00:34:13.880 --> 00:34:17.960 align:middle line:84% down to next year, I'll be downed at about $0.15-$0.20 00:34:17.960 --> 00:34:19.310 align:middle line:90% a SNP genotype. 00:34:19.310 --> 00:34:26.510 align:middle line:84% And there are three million SNPs minimum-- 00:34:26.510 --> 00:34:27.800 align:middle line:90% three to five million-- 00:34:27.800 --> 00:34:28.889 align:middle line:90% in the human genome. 00:34:28.889 --> 00:34:31.100 align:middle line:84% Now I'm not going to type at three million, 00:34:31.100 --> 00:34:32.947 align:middle line:90% but I've got to type-- 00:34:32.947 --> 00:34:34.489 align:middle line:84% in any one experiment, I've got to be 00:34:34.489 --> 00:34:42.995 align:middle line:84% able to type 1,000 over a 1020 megabits region-- 00:34:42.995 --> 00:34:44.120 align:middle line:90% one of these linkage peaks. 00:34:44.120 --> 00:34:46.489 align:middle line:84% So I got to do a lot of genotyping and a lot of people, 00:34:46.489 --> 00:34:48.050 align:middle line:90% and it's expensive. 00:34:48.050 --> 00:34:50.810 align:middle line:84% The very first position we cloned gene for asthma 00:34:50.810 --> 00:34:53.719 align:middle line:84% took six years and $15.6 million. 00:34:53.719 --> 00:34:57.200 align:middle line:84% We could do that experiment today for $2 million 00:34:57.200 --> 00:35:01.220 align:middle line:84% and a regular NIH grant, and that has totally 00:35:01.220 --> 00:35:02.210 align:middle line:90% changed the field. 00:35:02.210 --> 00:35:05.645 align:middle line:84% That's the kind of thing that's really making this possible. 00:35:05.645 --> 00:35:08.780 align:middle line:90% 00:35:08.780 --> 00:35:12.320 align:middle line:84% So I already said that if you're going 00:35:12.320 --> 00:35:14.510 align:middle line:84% to think like a geneticist, everybody 00:35:14.510 --> 00:35:16.680 align:middle line:84% has to know a little bit of population genetics. 00:35:16.680 --> 00:35:18.263 align:middle line:84% So you have to understand the concepts 00:35:18.263 --> 00:35:21.200 align:middle line:84% of linkage disequilibrium, drifts, 00:35:21.200 --> 00:35:25.490 align:middle line:90% natural selection, et cetera. 00:35:25.490 --> 00:35:27.123 align:middle line:84% Remind you of the fact that somebody 00:35:27.123 --> 00:35:28.790 align:middle line:84% asked the President of the United States 00:35:28.790 --> 00:35:31.910 align:middle line:84% whether he believed in evolution, and his answer was, 00:35:31.910 --> 00:35:34.970 align:middle line:90% the jury's still out. 00:35:34.970 --> 00:35:35.690 align:middle line:90% AUDIENCE: Really? 00:35:35.690 --> 00:35:38.562 align:middle line:84% SCOTT WEISS: Yeah, that's what he said. 00:35:38.562 --> 00:35:40.520 align:middle line:84% AUDIENCE: This is your former classmate, right? 00:35:40.520 --> 00:35:42.045 align:middle line:90% SCOTT WEISS: Right. 00:35:42.045 --> 00:35:43.670 align:middle line:84% I went to high school with a president. 00:35:43.670 --> 00:35:47.180 align:middle line:84% And so this is the first question 00:35:47.180 --> 00:35:50.600 align:middle line:84% that you're going to get asked if you're writing a grant. 00:35:50.600 --> 00:35:52.835 align:middle line:84% First thing that you have to address 00:35:52.835 --> 00:35:55.207 align:middle line:84% is the disease or the phenotype that you're 00:35:55.207 --> 00:35:56.540 align:middle line:90% interested in-- is it heritable? 00:35:56.540 --> 00:35:59.060 align:middle line:84% So there's lots of different ways to measure this. 00:35:59.060 --> 00:36:02.600 align:middle line:84% You can calculate a heritability estimates. 00:36:02.600 --> 00:36:03.950 align:middle line:90% You can do twin studies. 00:36:03.950 --> 00:36:05.840 align:middle line:84% You can develop this concept of risk 00:36:05.840 --> 00:36:11.510 align:middle line:84% to relatives, which is you look at the risk 00:36:11.510 --> 00:36:15.380 align:middle line:90% in the pro bands divided by-- 00:36:15.380 --> 00:36:17.930 align:middle line:84% the relatives divided by the risk in the population 00:36:17.930 --> 00:36:20.660 align:middle line:84% at large, or you can look at familial aggregation. 00:36:20.660 --> 00:36:23.370 align:middle line:84% But the point is, you've got to gather the evidence, 00:36:23.370 --> 00:36:25.805 align:middle line:84% and if you don't know that your phenotype is heritable, 00:36:25.805 --> 00:36:27.680 align:middle line:84% you're going to have to demonstrate that it's 00:36:27.680 --> 00:36:29.690 align:middle line:84% inheritable before anybody is going to give you a grant 00:36:29.690 --> 00:36:31.940 align:middle line:84% to study it because that's what geneticists say. 00:36:31.940 --> 00:36:33.890 align:middle line:90% They say they want to know that. 00:36:33.890 --> 00:36:35.820 align:middle line:84% They want to know the answer to that question. 00:36:35.820 --> 00:36:39.380 align:middle line:84% So I think it's-- asthma doesn't necessarily have a high 00:36:39.380 --> 00:36:42.170 align:middle line:84% heritability, but it clearly is a heritable disease. 00:36:42.170 --> 00:36:47.090 align:middle line:84% This is data from one twin study from Danish Twin Registry 00:36:47.090 --> 00:36:50.810 align:middle line:84% that looked at the concordance of asthma 00:36:50.810 --> 00:36:53.030 align:middle line:84% in identical and fraternal twins. 00:36:53.030 --> 00:36:56.570 align:middle line:84% Identical twins share 100% of their genotype. 00:36:56.570 --> 00:37:04.700 align:middle line:84% Fraternal twins share 50% of their alleles. 00:37:04.700 --> 00:37:08.340 align:middle line:84% Everybody knows the twins also share the environment. 00:37:08.340 --> 00:37:11.820 align:middle line:84% So that's another factor that's at issue here. 00:37:11.820 --> 00:37:14.060 align:middle line:84% But the reality is that there clearly 00:37:14.060 --> 00:37:16.280 align:middle line:84% is evidence of heritability of a disease. 00:37:16.280 --> 00:37:17.510 align:middle line:90% You get very different-- 00:37:17.510 --> 00:37:20.690 align:middle line:84% The problem here is that heritability estimates 00:37:20.690 --> 00:37:25.340 align:middle line:84% are always dependent on environmental exposures 00:37:25.340 --> 00:37:27.470 align:middle line:84% as well because the true underlying model-- 00:37:27.470 --> 00:37:30.290 align:middle line:84% and disease prevalence-- so the true underlying 00:37:30.290 --> 00:37:32.450 align:middle line:84% model for all of these diseases is clearly 00:37:32.450 --> 00:37:35.250 align:middle line:84% still going to be a gene by environment interaction. 00:37:35.250 --> 00:37:38.900 align:middle line:84% So after you've decided that the phenotypes that you're 00:37:38.900 --> 00:37:41.480 align:middle line:84% interested in are heritable, then you've got to go out 00:37:41.480 --> 00:37:43.760 align:middle line:84% and you've got to say, OK, I've got these phenotypes 00:37:43.760 --> 00:37:47.390 align:middle line:84% and I'm going to genotype them in a population. 00:37:47.390 --> 00:37:49.760 align:middle line:84% You can either look at disease phenotypes-- 00:37:49.760 --> 00:37:52.430 align:middle line:84% the advantages of this is that people want to look at asthma. 00:37:52.430 --> 00:37:53.960 align:middle line:90% They want asthma genes. 00:37:53.960 --> 00:37:57.710 align:middle line:84% They want to find, quote, the gene for asthma, unquote, 00:37:57.710 --> 00:38:01.310 align:middle line:84% which we already know is probably a false concept. 00:38:01.310 --> 00:38:03.680 align:middle line:84% But the problem with a lot of disease phenotypes 00:38:03.680 --> 00:38:07.790 align:middle line:84% is that even though there may be binary clinically, 00:38:07.790 --> 00:38:10.430 align:middle line:84% there may be real problems in terms 00:38:10.430 --> 00:38:13.370 align:middle line:84% of making that diagnosis in a way that would 00:38:13.370 --> 00:38:17.360 align:middle line:90% be useful for a research study. 00:38:17.360 --> 00:38:19.760 align:middle line:84% The problem with asthma is it's a syndrome. 00:38:19.760 --> 00:38:23.480 align:middle line:84% I mean, there is no one way of diagnosing asthma 00:38:23.480 --> 00:38:25.880 align:middle line:84% so that you can say, you take this test, 00:38:25.880 --> 00:38:27.710 align:middle line:84% and I can guarantee you that everybody 00:38:27.710 --> 00:38:30.420 align:middle line:84% that takes this test is going to have the disease 00:38:30.420 --> 00:38:33.650 align:middle line:84% and everybody who has a negative test doesn't have the disease. 00:38:33.650 --> 00:38:35.270 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:38:35.270 --> 00:38:40.030 align:middle line:90% 00:38:40.030 --> 00:38:42.340 align:middle line:84% SCOTT WEISS: Yeah, but the FEV1 doesn't tell you 00:38:42.340 --> 00:38:44.380 align:middle line:84% whether somebody's got asthma or not. 00:38:44.380 --> 00:38:48.220 align:middle line:84% I can show you people who have reduced FEV1 00:38:48.220 --> 00:38:51.550 align:middle line:84% and have cystic fibrosis or have interstitial lung disease 00:38:51.550 --> 00:38:53.590 align:middle line:90% or have COPD. 00:38:53.590 --> 00:38:55.870 align:middle line:84% I mean, they can have a lot of different things. 00:38:55.870 --> 00:38:59.410 align:middle line:84% So it lacks sensitivity and specificity the FEV1. 00:38:59.410 --> 00:39:01.420 align:middle line:84% And that's true for every single test. 00:39:01.420 --> 00:39:04.390 align:middle line:84% I mean, elevated IgE-- well, you could 00:39:04.390 --> 00:39:06.610 align:middle line:84% have elevated IgE from parasitic disease 00:39:06.610 --> 00:39:10.192 align:middle line:84% or from eosinophilic pneumonia or from 20 00:39:10.192 --> 00:39:11.150 align:middle line:90% other different things. 00:39:11.150 --> 00:39:17.290 align:middle line:84% So there is no single test, and the same 00:39:17.290 --> 00:39:21.370 align:middle line:84% may be true for most complex traits. 00:39:21.370 --> 00:39:23.230 align:middle line:84% There may be some phenotypes that 00:39:23.230 --> 00:39:27.080 align:middle line:84% are a little easier to measure like, say, well, 00:39:27.080 --> 00:39:29.450 align:middle line:90% I want to study obesity. 00:39:29.450 --> 00:39:32.530 align:middle line:90% Well, how fat is fat? 00:39:32.530 --> 00:39:36.730 align:middle line:84% Or is people who are fat like this different 00:39:36.730 --> 00:39:39.280 align:middle line:84% from people who are fat like this? 00:39:39.280 --> 00:39:42.160 align:middle line:84% I mean, there's all sorts of different ways 00:39:42.160 --> 00:39:45.070 align:middle line:90% of looking fat or being fat. 00:39:45.070 --> 00:39:49.930 align:middle line:84% So any one of these phenotypes has complications. 00:39:49.930 --> 00:39:53.800 align:middle line:90% And I can tell you this from-- 00:39:53.800 --> 00:39:56.800 align:middle line:84% when I first got into this, all I knew was phenotype. 00:39:56.800 --> 00:39:58.540 align:middle line:90% I was a world-class phenotyper. 00:39:58.540 --> 00:40:01.095 align:middle line:84% I knew all of the nuances of phenotype and everything 00:40:01.095 --> 00:40:02.470 align:middle line:84% there is to know about phenotype. 00:40:02.470 --> 00:40:04.210 align:middle line:84% And that tends to be what happens 00:40:04.210 --> 00:40:07.180 align:middle line:84% when you talk to clinicians because they understand 00:40:07.180 --> 00:40:07.780 align:middle line:90% that more. 00:40:07.780 --> 00:40:10.675 align:middle line:84% So this stuff is really, really important, 00:40:10.675 --> 00:40:12.550 align:middle line:84% but it's not going to get you very far if you 00:40:12.550 --> 00:40:14.050 align:middle line:90% don't know all the other stuff. 00:40:14.050 --> 00:40:15.760 align:middle line:84% You've got to know all the other things. 00:40:15.760 --> 00:40:19.030 align:middle line:84% I think the point is you do have to know this, 00:40:19.030 --> 00:40:19.822 align:middle line:90% and, again it's a-- 00:40:19.822 --> 00:40:22.030 align:middle line:84% AUDIENCE: The current problem is a lot of genomicists 00:40:22.030 --> 00:40:22.990 align:middle line:90% that don't understand. 00:40:22.990 --> 00:40:23.680 align:middle line:90% SCOTT WEISS: Exactly. 00:40:23.680 --> 00:40:24.850 align:middle line:84% Well, it goes back to the point that I 00:40:24.850 --> 00:40:26.225 align:middle line:84% was making earlier, which I think 00:40:26.225 --> 00:40:29.950 align:middle line:84% is that genetics is moving from a field 00:40:29.950 --> 00:40:37.360 align:middle line:84% where genetics were generalists to a field where geneticists 00:40:37.360 --> 00:40:38.320 align:middle line:90% are specialists. 00:40:38.320 --> 00:40:42.190 align:middle line:84% You get people who specialize in respiratory genetics, 00:40:42.190 --> 00:40:46.210 align:middle line:84% cardiovascular genetics, obesity genetics, diabetes genetics. 00:40:46.210 --> 00:40:48.910 align:middle line:84% The days of the person that can roam around 00:40:48.910 --> 00:40:50.290 align:middle line:90% and do all of these things-- 00:40:50.290 --> 00:40:53.260 align:middle line:84% no, I don't think that's going to happen. 00:40:53.260 --> 00:40:54.843 align:middle line:84% In five years, six years, you're going 00:40:54.843 --> 00:40:56.218 align:middle line:84% to have to be able to go in there 00:40:56.218 --> 00:40:58.360 align:middle line:84% and focus on a specific disease because it's 00:40:58.360 --> 00:41:02.245 align:middle line:84% going to be too complicated for you to be able to do otherwise. 00:41:02.245 --> 00:41:04.120 align:middle line:84% Then you've got this other type of phenotypes 00:41:04.120 --> 00:41:08.020 align:middle line:84% where you can say, well, OK, we want to look at asthma, 00:41:08.020 --> 00:41:11.320 align:middle line:84% but what about looking at intermediate phenotypes? 00:41:11.320 --> 00:41:14.650 align:middle line:84% So give me some examples of an intermediate phenotype related 00:41:14.650 --> 00:41:16.420 align:middle line:90% to my disease of interest? 00:41:16.420 --> 00:41:19.480 align:middle line:84% What would be an intermediate phenotype? 00:41:19.480 --> 00:41:21.860 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:41:21.860 --> 00:41:23.110 align:middle line:90% SCOTT WEISS: Absolutely, FEV1. 00:41:23.110 --> 00:41:23.610 align:middle line:90% What else? 00:41:23.610 --> 00:41:27.820 align:middle line:90% 00:41:27.820 --> 00:41:29.280 align:middle line:84% I had it up on a number of slides. 00:41:29.280 --> 00:41:31.220 align:middle line:90% So IgE level, right? 00:41:31.220 --> 00:41:32.530 align:middle line:90% It's a measure of allergy. 00:41:32.530 --> 00:41:39.070 align:middle line:84% Skin test reactivity, airways responsiveness, symptom score, 00:41:39.070 --> 00:41:43.085 align:middle line:84% sputum production, exhaled and out. 00:41:43.085 --> 00:41:44.710 align:middle line:84% I mean, the list goes on and on and on. 00:41:44.710 --> 00:41:46.670 align:middle line:90% You can create 100s. 00:41:46.670 --> 00:41:49.150 align:middle line:84% So for obesity, it could be looking 00:41:49.150 --> 00:41:55.600 align:middle line:84% at body mass index is the primary phenotype 00:41:55.600 --> 00:41:59.200 align:middle line:84% to define obesity, but then you could look at absolute fat mass 00:41:59.200 --> 00:42:04.990 align:middle line:84% or percent body fat, or waist to hip ratio, or insulin 00:42:04.990 --> 00:42:11.140 align:middle line:84% resistance, or do CT scans of somebody's 00:42:11.140 --> 00:42:13.613 align:middle line:90% abdominal fat deposition. 00:42:13.613 --> 00:42:15.280 align:middle line:84% I mean, there's a million different ways 00:42:15.280 --> 00:42:16.960 align:middle line:90% of potentially going at this. 00:42:16.960 --> 00:42:19.420 align:middle line:84% The advantage here is that sometimes these 00:42:19.420 --> 00:42:23.380 align:middle line:84% are more objective than a subjective, oh, it's asthma, 00:42:23.380 --> 00:42:25.150 align:middle line:90% it's not asthma. 00:42:25.150 --> 00:42:27.940 align:middle line:84% And it may be closer to the gene in the sense 00:42:27.940 --> 00:42:30.580 align:middle line:84% that you've got somebody's IgE level, 00:42:30.580 --> 00:42:34.120 align:middle line:84% you have some idea of genes that determine that. 00:42:34.120 --> 00:42:35.890 align:middle line:90% And it can be quantitative. 00:42:35.890 --> 00:42:41.680 align:middle line:84% You can do a different approach statistically 00:42:41.680 --> 00:42:45.160 align:middle line:84% to quantitative traits then you can 00:42:45.160 --> 00:42:48.170 align:middle line:84% use if you're looking at binary traits. 00:42:48.170 --> 00:42:50.530 align:middle line:84% AUDIENCE: [INAUDIBLE],, like if someone comes in 00:42:50.530 --> 00:42:54.310 align:middle line:84% and you look at them and say, OK, this person 00:42:54.310 --> 00:42:56.558 align:middle line:90% has a symptom of asthma. 00:42:56.558 --> 00:42:57.850 align:middle line:90% They have difficulty breathing. 00:42:57.850 --> 00:43:03.550 align:middle line:84% [INAUDIBLE] then you start delving into looking at 00:43:03.550 --> 00:43:06.040 align:middle line:84% SCOTT WEISS: All these intermediate phenotypes. 00:43:06.040 --> 00:43:09.670 align:middle line:84% AUDIENCE: And that correlates then with narrowing 00:43:09.670 --> 00:43:12.850 align:middle line:84% the diagnosis from, OK, you don't really have [INAUDIBLE],, 00:43:12.850 --> 00:43:16.030 align:middle line:84% you have this or you have this type of asthma or is that-- 00:43:16.030 --> 00:43:21.493 align:middle line:84% SCOTT WEISS: Well, it's the way I prefer to think about it. 00:43:21.493 --> 00:43:23.410 align:middle line:84% And I think it's probably a better way for you 00:43:23.410 --> 00:43:26.540 align:middle line:84% to think about it is that you've got to get away-- 00:43:26.540 --> 00:43:29.620 align:middle line:84% this is where thinking like a doctor and a clinician is bad. 00:43:29.620 --> 00:43:31.150 align:middle line:84% In the world of clinical medicine, 00:43:31.150 --> 00:43:32.950 align:middle line:90% it's just like religion. 00:43:32.950 --> 00:43:35.450 align:middle line:84% You either have the disease or you don't. 00:43:35.450 --> 00:43:37.990 align:middle line:84% There's no such thing as being a little bit pregnant. 00:43:37.990 --> 00:43:40.720 align:middle line:84% You're pregnant or you're not pregnant. 00:43:40.720 --> 00:43:44.980 align:middle line:84% You have to have bypass surgery or you don't. 00:43:44.980 --> 00:43:47.440 align:middle line:84% Clinicians live in a binary world. 00:43:47.440 --> 00:43:50.200 align:middle line:84% Real scientists live in the world 00:43:50.200 --> 00:43:53.110 align:middle line:90% of continuous distributions. 00:43:53.110 --> 00:43:57.340 align:middle line:90% So you can have-- 00:43:57.340 --> 00:43:58.810 align:middle line:90% when are you fat? 00:43:58.810 --> 00:44:04.900 align:middle line:84% Are you fat with a body mass index of 23, 24, 25 26. 00:44:04.900 --> 00:44:08.950 align:middle line:84% When do you have high blood pressure when it's 130 over 80? 00:44:08.950 --> 00:44:10.930 align:middle line:90% Or 140 over 90? 00:44:10.930 --> 00:44:15.110 align:middle line:90% Or when is that? 00:44:15.110 --> 00:44:18.100 align:middle line:84% And the other thing is that way to think about these 00:44:18.100 --> 00:44:20.110 align:middle line:84% is kind of like overlapping Venn diagrams. 00:44:20.110 --> 00:44:22.690 align:middle line:84% The clinical phenotype is actually 00:44:22.690 --> 00:44:25.900 align:middle line:84% a composite of these overlapping Venn diagrams 00:44:25.900 --> 00:44:31.840 align:middle line:84% that all have separate genetic determinants and things that 00:44:31.840 --> 00:44:34.360 align:middle line:84% contribute to them, separate genetic-- 00:44:34.360 --> 00:44:37.300 align:middle line:84% it's like dissecting a layer and peeling an onion, 00:44:37.300 --> 00:44:41.360 align:middle line:84% where you've got all these different things. 00:44:41.360 --> 00:44:50.020 align:middle line:84% But I think, in many ways, being a clinician can help you 00:44:50.020 --> 00:44:51.070 align:middle line:90% as a research scientist. 00:44:51.070 --> 00:44:52.487 align:middle line:84% But in some ways, it can also hurt 00:44:52.487 --> 00:44:54.873 align:middle line:84% because you start to think in these absolute terms. 00:44:54.873 --> 00:44:56.290 align:middle line:84% So I think the better way to think 00:44:56.290 --> 00:45:01.630 align:middle line:84% about it is that these intermediate phenotypes overlap 00:45:01.630 --> 00:45:03.880 align:middle line:90% to create clinical phenotypes. 00:45:03.880 --> 00:45:06.820 align:middle line:84% And yes, what you're trying to do 00:45:06.820 --> 00:45:12.700 align:middle line:84% is stratify in some way or classify people in some way 00:45:12.700 --> 00:45:17.170 align:middle line:84% so you're creating homogeneity so that you can actually 00:45:17.170 --> 00:45:21.790 align:middle line:84% identify the genetic determinants of a disease 00:45:21.790 --> 00:45:23.180 align:middle line:90% or an intermediate phenotype. 00:45:23.180 --> 00:45:25.620 align:middle line:84% So you want to go in that direction. 00:45:25.620 --> 00:45:29.680 align:middle line:84% But most of these things lack sufficient sensitivity 00:45:29.680 --> 00:45:34.940 align:middle line:84% and specificity to really be terribly helpful. 00:45:34.940 --> 00:45:37.060 align:middle line:84% So this is just a list of some of the phenotypes 00:45:37.060 --> 00:45:39.255 align:middle line:84% that people have looked at in asthma. 00:45:39.255 --> 00:45:40.630 align:middle line:84% And I've studied some of the ones 00:45:40.630 --> 00:45:45.310 align:middle line:84% that people have focused on in terms of linkage peaks 00:45:45.310 --> 00:45:46.960 align:middle line:84% that have actually been identified. 00:45:46.960 --> 00:45:50.030 align:middle line:90% 00:45:50.030 --> 00:45:52.613 align:middle line:84% But this is interesting because there's clearly 00:45:52.613 --> 00:45:54.280 align:middle line:84% a bias in the literature because there's 00:45:54.280 --> 00:45:57.250 align:middle line:84% a whole bunch of these other phenotypes where 00:45:57.250 --> 00:45:58.210 align:middle line:90% you could just as-- 00:45:58.210 --> 00:46:00.850 align:middle line:84% and I could create a list of 30 more of these 00:46:00.850 --> 00:46:02.440 align:middle line:90% where people haven't looked at. 00:46:02.440 --> 00:46:04.660 align:middle line:84% So this just gets to the point that there's 00:46:04.660 --> 00:46:06.910 align:middle line:84% plenty of work here for anybody who 00:46:06.910 --> 00:46:10.625 align:middle line:84% wants to do this stuff because you can go out 00:46:10.625 --> 00:46:12.250 align:middle line:84% and I've got a junior person in my lab, 00:46:12.250 --> 00:46:14.833 align:middle line:84% he's got a bunch of phenotypes that he's really interested in, 00:46:14.833 --> 00:46:16.208 align:middle line:84% and he's going to go out and he's 00:46:16.208 --> 00:46:17.791 align:middle line:84% going to determine their heritability, 00:46:17.791 --> 00:46:19.583 align:middle line:84% and then he's going to write another grant, 00:46:19.583 --> 00:46:22.300 align:middle line:84% and he's going to map the genes for them and so on and so forth 00:46:22.300 --> 00:46:28.080 align:middle line:84% because he wants to have his own little area to work on. 00:46:28.080 --> 00:46:31.260 align:middle line:84% So then the next thing-- so now we're kind of at the point 00:46:31.260 --> 00:46:32.993 align:middle line:84% where you've got to I've got to move 00:46:32.993 --> 00:46:34.410 align:middle line:84% a little faster or we're not going 00:46:34.410 --> 00:46:35.618 align:middle line:90% to make our way through this. 00:46:35.618 --> 00:46:37.290 align:middle line:84% But you've got to have a study design, 00:46:37.290 --> 00:46:41.490 align:middle line:84% and there's a bunch of different ways of doing this. 00:46:41.490 --> 00:46:42.840 align:middle line:90% You can do linkage. 00:46:42.840 --> 00:46:45.030 align:middle line:90% You can do association. 00:46:45.030 --> 00:46:47.760 align:middle line:84% And amongst the linkage studies, you 00:46:47.760 --> 00:46:54.150 align:middle line:84% can do a little sharing methods which are distribution free, 00:46:54.150 --> 00:47:01.830 align:middle line:84% or you can do continuous distributions 00:47:01.830 --> 00:47:04.140 align:middle line:90% and focus on that. 00:47:04.140 --> 00:47:06.090 align:middle line:84% There are two types of genetic association 00:47:06.090 --> 00:47:08.700 align:middle line:84% studies-- the family-based and the case control. 00:47:08.700 --> 00:47:11.700 align:middle line:84% Important point here is that they're very different. 00:47:11.700 --> 00:47:16.200 align:middle line:84% Here, you have to genotype three people. 00:47:16.200 --> 00:47:19.230 align:middle line:84% Here, you have to genotype only two people. 00:47:19.230 --> 00:47:22.200 align:middle line:84% Different hypotheses-- here, you're 00:47:22.200 --> 00:47:27.480 align:middle line:84% looking at the alleles or the genotypes in the cases 00:47:27.480 --> 00:47:28.740 align:middle line:90% relative to controls. 00:47:28.740 --> 00:47:30.690 align:middle line:84% It's the frequency of the genotype, frequency 00:47:30.690 --> 00:47:32.880 align:middle line:84% in the cases, versus the controls here. 00:47:32.880 --> 00:47:37.650 align:middle line:84% You're looking at a transmitted alleles 00:47:37.650 --> 00:47:41.200 align:middle line:84% from a heterozygous parent to an affected offspring. 00:47:41.200 --> 00:47:44.850 align:middle line:84% So very different hypotheses, different study designs, 00:47:44.850 --> 00:47:47.850 align:middle line:84% and important thing to recognize is 00:47:47.850 --> 00:47:53.010 align:middle line:84% that in any association study, the association 00:47:53.010 --> 00:47:55.740 align:middle line:84% between a variant and a phenotype 00:47:55.740 --> 00:47:58.320 align:middle line:84% can be due to a causal relationship. 00:47:58.320 --> 00:48:01.170 align:middle line:84% It can be the linkage disequilibrium. 00:48:01.170 --> 00:48:03.420 align:middle line:84% Or it can be due to population admixture, which 00:48:03.420 --> 00:48:06.570 align:middle line:84% means that, usually in the context of the case control 00:48:06.570 --> 00:48:09.450 align:middle line:84% study, not a family-based study, you've got different allele 00:48:09.450 --> 00:48:12.570 align:middle line:84% frequencies segregating in the cases in the controls 00:48:12.570 --> 00:48:15.180 align:middle line:84% because you've got different population 00:48:15.180 --> 00:48:20.790 align:middle line:84% histories, evolutionary histories, that have determined 00:48:20.790 --> 00:48:22.180 align:middle line:90% those allele frequencies. 00:48:22.180 --> 00:48:24.780 align:middle line:84% So most extreme example would be I 00:48:24.780 --> 00:48:28.500 align:middle line:84% had 1,000 Italian cases of asthma, 00:48:28.500 --> 00:48:34.290 align:middle line:84% and I'm comparing it to 1,000 Swiss controls who 00:48:34.290 --> 00:48:35.370 align:middle line:90% don't have asthma. 00:48:35.370 --> 00:48:41.490 align:middle line:84% And even though these two groups are predominantly Caucasian, 00:48:41.490 --> 00:48:43.470 align:middle line:84% their evolutionary history may be different, 00:48:43.470 --> 00:48:44.880 align:middle line:84% and the allele frequencies may be 00:48:44.880 --> 00:48:46.390 align:middle line:90% different as a result of that. 00:48:46.390 --> 00:48:50.928 align:middle line:84% So even within an ethnic group, you 00:48:50.928 --> 00:48:52.720 align:middle line:84% can get these different allele frequencies. 00:48:52.720 --> 00:48:58.380 align:middle line:84% And this is because ethnicity or self-designated ethnicity 00:48:58.380 --> 00:49:05.400 align:middle line:84% is only a weak predictor of evolutionary history. 00:49:05.400 --> 00:49:07.750 align:middle line:84% AUDIENCE: What was your example or [INAUDIBLE] example. 00:49:07.750 --> 00:49:13.560 align:middle line:84% So you compare [INAUDIBLE] assoiciations [INAUDIBLE] 00:49:13.560 --> 00:49:18.900 align:middle line:84% between Germans and Italians, those two populations. 00:49:18.900 --> 00:49:22.320 align:middle line:84% Sure enough, we found a linkage association 00:49:22.320 --> 00:49:26.670 align:middle line:84% between pasta eating and some piece of genome 00:49:26.670 --> 00:49:28.560 align:middle line:84% because, in fact, what you'd be looking 00:49:28.560 --> 00:49:33.270 align:middle line:84% for is linkage to the fact that you're an Italian, just 00:49:33.270 --> 00:49:36.000 align:middle line:84% by the fact that Italians have a distinct [INAUDIBLE] 00:49:36.000 --> 00:49:38.010 align:middle line:84% polymorphisms then the Germans [INAUDIBLE] 00:49:38.010 --> 00:49:40.350 align:middle line:84% creates this [? past ?] association when, 00:49:40.350 --> 00:49:44.125 align:middle line:84% in fact, [INAUDIBLE] looking at different populations. 00:49:44.125 --> 00:49:45.750 align:middle line:84% SCOTT WEISS: Some of the guys in my lab 00:49:45.750 --> 00:49:47.970 align:middle line:84% wrote an article demonstrating all 00:49:47.970 --> 00:49:51.360 align:middle line:84% of the potential problems in the case control type 00:49:51.360 --> 00:49:54.342 align:middle line:90% of genetic association study. 00:49:54.342 --> 00:49:55.800 align:middle line:84% And one of the things that's really 00:49:55.800 --> 00:49:57.870 align:middle line:84% impressive about this paper in Science 00:49:57.870 --> 00:50:01.380 align:middle line:84% is that we all use genetic association 00:50:01.380 --> 00:50:05.700 align:middle line:84% as part of the fine mapping process to map a linkage peak. 00:50:05.700 --> 00:50:10.650 align:middle line:84% But-- this is very important-- but 00:50:10.650 --> 00:50:14.340 align:middle line:84% because even if you can get rid of the population admixture 00:50:14.340 --> 00:50:18.215 align:middle line:84% problem, linkage disequilibrium is always an issue. 00:50:18.215 --> 00:50:19.590 align:middle line:84% And so you're never going to know 00:50:19.590 --> 00:50:25.900 align:middle line:84% for sure if you're at the gene or you're just close by to it. 00:50:25.900 --> 00:50:27.990 align:middle line:84% And so you're going to have to have 00:50:27.990 --> 00:50:32.190 align:middle line:84% something else to show that you've actually found the gene. 00:50:32.190 --> 00:50:36.480 align:middle line:84% You're not getting into science just with genetic association. 00:50:36.480 --> 00:50:39.690 align:middle line:84% And so the people in the paper that's coming out this week, 00:50:39.690 --> 00:50:46.590 align:middle line:84% they have expressed the gene in bronchial tissue. 00:50:46.590 --> 00:50:51.900 align:middle line:84% They've done immunohistochemistry 00:50:51.900 --> 00:50:55.320 align:middle line:84% to show that the gene is expressed in epithelium. 00:50:55.320 --> 00:50:58.500 align:middle line:84% They replicated their results in a different population, 00:50:58.500 --> 00:51:00.300 align:middle line:90% et cetera. 00:51:00.300 --> 00:51:04.100 align:middle line:84% So the thing about the [INAUDIBLE] 00:51:04.100 --> 00:51:06.270 align:middle line:84% about the case control studies and about even 00:51:06.270 --> 00:51:09.570 align:middle line:84% family-based association is that these studies are really 00:51:09.570 --> 00:51:10.780 align:middle line:90% easy to do. 00:51:10.780 --> 00:51:13.000 align:middle line:84% And so there's lots of them in the literature. 00:51:13.000 --> 00:51:15.200 align:middle line:84% So it's really important for you to know, 00:51:15.200 --> 00:51:17.610 align:middle line:84% going back to this slide, it's really important for you 00:51:17.610 --> 00:51:23.557 align:middle line:84% to know these potential problems because you want to be able 00:51:23.557 --> 00:51:25.140 align:middle line:84% to read this literature and say, yeah, 00:51:25.140 --> 00:51:28.120 align:middle line:84% these guys really found something or maybe they didn't. 00:51:28.120 --> 00:51:32.580 align:middle line:84% So the advantages of this candidate gene thing 00:51:32.580 --> 00:51:38.730 align:middle line:84% is that it's cheap and easy to compare-- remember, 00:51:38.730 --> 00:51:42.540 align:middle line:84% I said that now, four positionally cloned genes that 00:51:42.540 --> 00:51:51.060 align:middle line:84% have used this type of genome screen approach, four that 00:51:51.060 --> 00:51:53.370 align:middle line:84% have been identified since the human genome 00:51:53.370 --> 00:51:54.840 align:middle line:90% was mapped in 1996. 00:51:54.840 --> 00:51:56.740 align:middle line:90% Well, that's seven years. 00:51:56.740 --> 00:51:58.170 align:middle line:90% That's not even one gene a year. 00:51:58.170 --> 00:52:02.080 align:middle line:90% That's pretty meek, or weak. 00:52:02.080 --> 00:52:05.730 align:middle line:84% And that's because this is very expensive, technologically 00:52:05.730 --> 00:52:06.550 align:middle line:90% intensive. 00:52:06.550 --> 00:52:08.670 align:middle line:84% But the thing that's great about this 00:52:08.670 --> 00:52:11.610 align:middle line:84% is come up with a novel gene at the end of the time. 00:52:11.610 --> 00:52:14.100 align:middle line:84% So it's not dependent on what anybody 00:52:14.100 --> 00:52:15.440 align:middle line:90% knows about pathobiology. 00:52:15.440 --> 00:52:17.100 align:middle line:84% And so you can go this way, and you 00:52:17.100 --> 00:52:20.530 align:middle line:84% can say, look, I know that IgE is important in asthma. 00:52:20.530 --> 00:52:25.110 align:middle line:84% So I know that we ought to be screening IL-13 IL-4, 00:52:25.110 --> 00:52:32.010 align:middle line:84% IL-4 alpha receptor, CTLA-4, all of those genes in the pathway 00:52:32.010 --> 00:52:34.080 align:middle line:84% that determines IgE makes sense-- 00:52:34.080 --> 00:52:36.090 align:middle line:84% screen those genes because we've already 00:52:36.090 --> 00:52:38.040 align:middle line:84% said that people with asthma have high IgE. 00:52:38.040 --> 00:52:42.000 align:middle line:84% Well, you check those genes and yeah, in fact, they are-- 00:52:42.000 --> 00:52:47.760 align:middle line:84% most of those genes are asthma or allergy genes. 00:52:47.760 --> 00:52:49.800 align:middle line:90% It's not real exciting, though. 00:52:49.800 --> 00:52:52.920 align:middle line:84% I mean, it's not like everybody's going to jump up 00:52:52.920 --> 00:52:56.370 align:middle line:84% and say, oh my god, IL-13 is an asthma gene. 00:52:56.370 --> 00:52:58.590 align:middle line:84% Well, molecular biologist says, yeah, well, we 00:52:58.590 --> 00:52:59.790 align:middle line:90% knew that 10 years ago. 00:52:59.790 --> 00:53:01.350 align:middle line:90% What's new? 00:53:01.350 --> 00:53:04.230 align:middle line:90% What's great about that? 00:53:04.230 --> 00:53:06.518 align:middle line:84% I mean, there are interesting things about it 00:53:06.518 --> 00:53:08.310 align:middle line:84% because you actually can get to the level-- 00:53:08.310 --> 00:53:10.320 align:middle line:84% it's going to change molecular biology too 00:53:10.320 --> 00:53:12.130 align:middle line:84% because you're actually going to get the level where you say, 00:53:12.130 --> 00:53:14.400 align:middle line:84% well, it's these three variants in the promoter, 00:53:14.400 --> 00:53:17.040 align:middle line:84% it's this variant in exon 1, and it's 00:53:17.040 --> 00:53:19.710 align:middle line:84% this particular haplotype that's determining 00:53:19.710 --> 00:53:21.570 align:middle line:90% the effect on IgE level. 00:53:21.570 --> 00:53:24.368 align:middle line:84% So molecular biology is going to change because people 00:53:24.368 --> 00:53:25.410 align:middle line:90% aren't going to just be-- 00:53:25.410 --> 00:53:27.870 align:middle line:84% aren't going to get away with knocking out a gene 00:53:27.870 --> 00:53:30.225 align:middle line:84% or looking at a whole gene effect. 00:53:30.225 --> 00:53:32.100 align:middle line:84% They're going to actually have to go in there 00:53:32.100 --> 00:53:35.790 align:middle line:84% and determine the particular variants that 00:53:35.790 --> 00:53:39.250 align:middle line:84% are important in terms of the molecular mechanisms. 00:53:39.250 --> 00:53:41.940 align:middle line:84% So I don't want to denigrate this 00:53:41.940 --> 00:53:43.860 align:middle line:84% because this is-- we do all of us-- 00:53:43.860 --> 00:53:46.968 align:middle line:84% do a lot of this stuff to keep ourselves busy while we're 00:53:46.968 --> 00:53:49.260 align:middle line:84% trying to do these really big experiments that are very 00:53:49.260 --> 00:53:52.080 align:middle line:90% expensive and take a long time. 00:53:52.080 --> 00:53:52.780 align:middle line:90% Skip that. 00:53:52.780 --> 00:53:54.870 align:middle line:84% So let's talk a little bit about linkage. 00:53:54.870 --> 00:53:58.290 align:middle line:84% Linkage is this idea of take these microsatellite markers 00:53:58.290 --> 00:54:00.970 align:middle line:90% all the way across the genome. 00:54:00.970 --> 00:54:02.490 align:middle line:90% It's a property of families. 00:54:02.490 --> 00:54:04.320 align:middle line:84% It's not a property of individuals. 00:54:04.320 --> 00:54:05.820 align:middle line:84% And you're looking to see if there's 00:54:05.820 --> 00:54:08.310 align:middle line:84% a particular region of the genome contains a gene that's 00:54:08.310 --> 00:54:10.380 align:middle line:84% related to the phenotype of interest that's 00:54:10.380 --> 00:54:19.360 align:middle line:84% segregating in these families, using identity by descent. 00:54:19.360 --> 00:54:23.160 align:middle line:84% So what you do is, when you have some extended pedigree 00:54:23.160 --> 00:54:27.510 align:middle line:84% like this, what you could do is you 00:54:27.510 --> 00:54:32.550 align:middle line:84% could do segregation analysis to develop a model 00:54:32.550 --> 00:54:34.560 align:middle line:84% to see how the disease is actually 00:54:34.560 --> 00:54:36.240 align:middle line:90% segregating in this population. 00:54:36.240 --> 00:54:39.240 align:middle line:84% But that's pretty difficult for complex traits. 00:54:39.240 --> 00:54:40.335 align:middle line:90% It's not easy to do. 00:54:40.335 --> 00:54:43.240 align:middle line:90% 00:54:43.240 --> 00:54:46.043 align:middle line:84% You could also use this approach, 00:54:46.043 --> 00:54:47.460 align:middle line:84% the allele-sharing approach, which 00:54:47.460 --> 00:54:49.800 align:middle line:90% assumes no mode of inheritance. 00:54:49.800 --> 00:54:53.040 align:middle line:84% It just says we collected a whole bunch of sib pairs 00:54:53.040 --> 00:54:54.660 align:middle line:84% who are affected, and we're going 00:54:54.660 --> 00:54:57.570 align:middle line:84% to test whether these affected relatives have inherited 00:54:57.570 --> 00:55:00.900 align:middle line:84% a region of the genome identity by descent 00:55:00.900 --> 00:55:04.350 align:middle line:84% more often than expected under a random Mendelian segregation. 00:55:04.350 --> 00:55:07.420 align:middle line:84% And the nice thing about this is that it's easy, 00:55:07.420 --> 00:55:09.030 align:middle line:90% but it's not very powerful. 00:55:09.030 --> 00:55:12.540 align:middle line:84% I mean, the problem is you need a lot of sib pairs 00:55:12.540 --> 00:55:15.690 align:middle line:84% and even then even with over 300 sib pairs, 00:55:15.690 --> 00:55:19.690 align:middle line:84% you don't get such great power using this approach. 00:55:19.690 --> 00:55:27.390 align:middle line:84% So power goes up if the disease is more heritable, 00:55:27.390 --> 00:55:29.100 align:middle line:84% and you can do with less sib pairs, 00:55:29.100 --> 00:55:32.220 align:middle line:84% but the reality is that even with a huge number of sib 00:55:32.220 --> 00:55:34.770 align:middle line:84% pairs, you may not have a lot of power 00:55:34.770 --> 00:55:39.900 align:middle line:84% if the lambda is down here, which it is for asthma, 00:55:39.900 --> 00:55:40.440 align:middle line:90% probably. 00:55:40.440 --> 00:55:45.090 align:middle line:84% So I think that this is why people have focused 00:55:45.090 --> 00:55:49.170 align:middle line:84% on extended pedigrees in these relative genetic isolates, 00:55:49.170 --> 00:55:52.830 align:middle line:84% and that's why we're so excited about Costa Rica. 00:55:52.830 --> 00:55:55.620 align:middle line:84% The Finns are clearly excited about Finland. 00:55:55.620 --> 00:55:59.647 align:middle line:84% And Decode is doing what it's doing in Iceland. 00:55:59.647 --> 00:56:01.480 align:middle line:84% Whether we're going to be successful or not, 00:56:01.480 --> 00:56:02.022 align:middle line:90% I don't know. 00:56:02.022 --> 00:56:05.070 align:middle line:84% But the basic approach is that whether you're 00:56:05.070 --> 00:56:08.400 align:middle line:84% using an outbred population or a genetic isolate, 00:56:08.400 --> 00:56:11.160 align:middle line:84% and whether you're using sib pairs or pedigrees, 00:56:11.160 --> 00:56:17.550 align:middle line:84% is you've got these usually di- and trinucleotide repeat STR 00:56:17.550 --> 00:56:22.560 align:middle line:84% microsatellite markers, most of the genome services 00:56:22.560 --> 00:56:26.860 align:middle line:84% use about 400 of these markers equally randomly spaced 00:56:26.860 --> 00:56:28.720 align:middle line:84% across the genome, and what you do 00:56:28.720 --> 00:56:32.590 align:middle line:84% is do just do a form of logistic regression 00:56:32.590 --> 00:56:36.670 align:middle line:84% basically where you do a LOD score, log of the odds ratio 00:56:36.670 --> 00:56:39.760 align:middle line:84% calculation between relating phenotype 00:56:39.760 --> 00:56:42.790 align:middle line:84% in the family to these markers, and what 00:56:42.790 --> 00:56:46.150 align:middle line:84% you do is get a linkage peak that is the log 00:56:46.150 --> 00:56:49.600 align:middle line:84% score for that relationship between the markers 00:56:49.600 --> 00:56:53.350 align:middle line:90% and the phenotype. 00:56:53.350 --> 00:56:58.870 align:middle line:84% And what that says is there's a gene or multiple genes 00:56:58.870 --> 00:57:02.680 align:middle line:84% in this particular region on a chromosome that's associated 00:57:02.680 --> 00:57:04.092 align:middle line:90% with a particular phenotype. 00:57:04.092 --> 00:57:06.550 align:middle line:84% And then you have to then go in down and put more markers-- 00:57:06.550 --> 00:57:10.330 align:middle line:84% first more STR markers and then SNPs-- 00:57:10.330 --> 00:57:13.360 align:middle line:84% and gradually map that region until you've actually 00:57:13.360 --> 00:57:21.040 align:middle line:84% got it down to a very small region of a particular 1,000 00:57:21.040 --> 00:57:22.570 align:middle line:90% base pairs or whatever. 00:57:22.570 --> 00:57:27.760 align:middle line:84% We can say it's a gene or one or two genes in this relatively 00:57:27.760 --> 00:57:28.600 align:middle line:90% large region. 00:57:28.600 --> 00:57:31.520 align:middle line:84% So that takes a lot of genotyping and a lot of work. 00:57:31.520 --> 00:57:34.390 align:middle line:84% So our experiments, now, over the next year, 00:57:34.390 --> 00:57:37.390 align:middle line:84% we have all these linkage peaks in asthma and COPD. 00:57:37.390 --> 00:57:39.850 align:middle line:84% Each experiment is going to be about $200,000. 00:57:39.850 --> 00:57:43.700 align:middle line:84% There's going to be 1,500-1,600 SNPs in each of these regions, 00:57:43.700 --> 00:57:47.140 align:middle line:84% and we're going to fine map three or four regions over 00:57:47.140 --> 00:57:48.400 align:middle line:90% the course of the next year. 00:57:48.400 --> 00:57:51.400 align:middle line:84% And hopefully, we will be in Science. 00:57:51.400 --> 00:57:53.232 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:57:53.232 --> 00:57:59.500 align:middle line:90% 00:57:59.500 --> 00:58:05.620 align:middle line:84% SCOTT WEISS: SNPs 400 base pairs kind of thing, 00:58:05.620 --> 00:58:11.590 align:middle line:84% on the order of one per 1,000-- about one per 1,000 bases. 00:58:11.590 --> 00:58:14.090 align:middle line:84% That's about what we're shooting for. 00:58:14.090 --> 00:58:16.570 align:middle line:84% So this is just a summary of all of the genome screens that 00:58:16.570 --> 00:58:20.290 align:middle line:84% have been done in asthma just to show you that most of them 00:58:20.290 --> 00:58:22.360 align:middle line:90% have been sib pair studies. 00:58:22.360 --> 00:58:24.460 align:middle line:84% Most of them have been relatively small. 00:58:24.460 --> 00:58:27.220 align:middle line:84% But we do get a substantial amount of replication. 00:58:27.220 --> 00:58:29.410 align:middle line:84% These are regions across the genome-- this one right 00:58:29.410 --> 00:58:31.510 align:middle line:84% here, that's the gene that was just mapped. 00:58:31.510 --> 00:58:34.150 align:middle line:90% 00:58:34.150 --> 00:58:36.760 align:middle line:84% Several populations, including the Finns, 00:58:36.760 --> 00:58:39.040 align:middle line:90% showed a peak in this region. 00:58:39.040 --> 00:58:45.400 align:middle line:84% And they got this gene, and then they went to the Canadians 00:58:45.400 --> 00:58:48.370 align:middle line:84% and they said, can we replicate it in your population? 00:58:48.370 --> 00:58:52.840 align:middle line:84% The interesting thing is it was asthma in the Finns, 00:58:52.840 --> 00:58:55.060 align:middle line:84% but it's high IgE in the Canadians. 00:58:55.060 --> 00:58:58.780 align:middle line:84% So it shows you that this problem 00:58:58.780 --> 00:59:02.380 align:middle line:84% of phenotypic heterogeneity and genetic heterogeneity 00:59:02.380 --> 00:59:04.990 align:middle line:90% is a big issue here. 00:59:04.990 --> 00:59:07.810 align:middle line:84% So it isn't a perfect replication 00:59:07.810 --> 00:59:10.450 align:middle line:84% at the phenotype level between these two populations, 00:59:10.450 --> 00:59:12.850 align:middle line:84% but they've got all this other stuff-- the expression 00:59:12.850 --> 00:59:15.820 align:middle line:84% and everything else-- that proves that they've really 00:59:15.820 --> 00:59:17.230 align:middle line:90% got the gene. 00:59:17.230 --> 00:59:23.260 align:middle line:84% But the one we're working on is actually not on here. 00:59:23.260 --> 00:59:27.280 align:middle line:84% I didn't leave it off intentionally, but it's 12q, 00:59:27.280 --> 00:59:30.200 align:middle line:84% and it's one of the ones that's the most replicable. 00:59:30.200 --> 00:59:32.120 align:middle line:90% Now here it is here. 00:59:32.120 --> 00:59:34.930 align:middle line:90% It's in this slide right here. 00:59:34.930 --> 00:59:38.080 align:middle line:84% So this is a very good region, and there's an [INAUDIBLE] 00:59:38.080 --> 00:59:41.300 align:middle line:84% here, but it's also got a very low p-value. 00:59:41.300 --> 00:59:42.910 align:middle line:84% So that's one of the better ones. 00:59:42.910 --> 00:59:45.460 align:middle line:84% Now you can already see from this, 00:59:45.460 --> 00:59:48.160 align:middle line:84% each one of these-- this region has five or six different genes 00:59:48.160 --> 00:59:49.120 align:middle line:90% in this region. 00:59:49.120 --> 00:59:52.600 align:middle line:84% It's the cytokine clusters here, beta 2 adrenergic receptors 00:59:52.600 --> 00:59:55.280 align:middle line:84% here, IL-13 is here, CD-14 is here. 00:59:55.280 --> 00:59:57.430 align:middle line:84% So there's a whole bunch of small genes in here. 00:59:57.430 --> 01:00:00.530 align:middle line:84% Nobody knows whether there's a big gene or not. 01:00:00.530 --> 01:00:04.000 align:middle line:84% And it may be that linkage peak is just being given by the fact 01:00:04.000 --> 01:00:08.710 align:middle line:84% that there's a whole bunch of small genes in that region. 01:00:08.710 --> 01:00:13.510 align:middle line:84% This one, the one we're working in, this is 30 mega bases. 01:00:13.510 --> 01:00:16.600 align:middle line:90% That's huge-- huge region. 01:00:16.600 --> 01:00:19.120 align:middle line:84% But you can see from just looking at this that-- 01:00:19.120 --> 01:00:21.925 align:middle line:84% one, two, three, four, five, six, seven, eight, nine-- 01:00:21.925 --> 01:00:22.800 align:middle line:90% and there's another-- 01:00:22.800 --> 01:00:24.640 align:middle line:84% I mean, these are 20 regions, each of them 01:00:24.640 --> 01:00:30.700 align:middle line:90% about 20 to 40 mega bases. 01:00:30.700 --> 01:00:33.490 align:middle line:84% There could be five or six genes in each one of these regions, 01:00:33.490 --> 01:00:36.040 align:middle line:84% and at least two of the positionally clonal genes-- 01:00:36.040 --> 01:00:38.680 align:middle line:84% there were two genes in the region, 01:00:38.680 --> 01:00:41.140 align:middle line:84% and you couldn't tell from the articles. 01:00:41.140 --> 01:00:43.690 align:middle line:84% In fact, this Finnish article that's about to come out-- 01:00:43.690 --> 01:00:45.460 align:middle line:84% there's a second gene identified, 01:00:45.460 --> 01:00:47.770 align:middle line:84% and they don't have the molecular biology 01:00:47.770 --> 01:00:49.400 align:middle line:90% on that in the paper. 01:00:49.400 --> 01:00:52.823 align:middle line:84% And they're not sure what that gene is doing. 01:00:52.823 --> 01:00:54.240 align:middle line:84% AUDIENCE: So you're actually going 01:00:54.240 --> 01:00:56.890 align:middle line:90% to do a [INAUDIBLE] genome. 01:00:56.890 --> 01:00:59.633 align:middle line:90% 01:00:59.633 --> 01:01:00.550 align:middle line:90% SCOTT WEISS: Probably. 01:01:00.550 --> 01:01:02.887 align:middle line:90% AUDIENCE: [INAUDIBLE] 01:01:02.887 --> 01:01:04.720 align:middle line:84% SCOTT WEISS: So these are some of the issues 01:01:04.720 --> 01:01:06.930 align:middle line:84% in doing the type of linkage studies 01:01:06.930 --> 01:01:09.430 align:middle line:84% that I talked about-- multiple markers, multiple phenotypes, 01:01:09.430 --> 01:01:10.990 align:middle line:90% multiple comparisons. 01:01:10.990 --> 01:01:13.540 align:middle line:84% Phenotypes are correlated, markers not independent-- 01:01:13.540 --> 01:01:14.663 align:middle line:90% you know you've got to do. 01:01:14.663 --> 01:01:16.330 align:middle line:84% So there's a lot of statistical issues-- 01:01:16.330 --> 01:01:19.300 align:middle line:84% this work is really exciting, I think, 01:01:19.300 --> 01:01:24.310 align:middle line:84% because it combines genetics, clinical medicine, 01:01:24.310 --> 01:01:27.933 align:middle line:84% molecular genetics, statistics, evolutionary-- 01:01:27.933 --> 01:01:29.600 align:middle line:84% all of this stuff is all mixed together. 01:01:29.600 --> 01:01:31.930 align:middle line:84% So a lot of important statistical issues 01:01:31.930 --> 01:01:33.880 align:middle line:90% in doing these genome screens. 01:01:33.880 --> 01:01:36.350 align:middle line:84% So then you got to genotype the people. 01:01:36.350 --> 01:01:41.537 align:middle line:84% We've already said that SNPs are the primary genetic variation 01:01:41.537 --> 01:01:42.370 align:middle line:90% in the human genome. 01:01:42.370 --> 01:01:47.200 align:middle line:84% But we found INDELs, we found repeats, 01:01:47.200 --> 01:01:52.360 align:middle line:84% we found SNPs and INDELs together. 01:01:52.360 --> 01:01:54.310 align:middle line:84% I mean, there's all kinds of stuff. 01:01:54.310 --> 01:01:58.000 align:middle line:84% In general, SNPs occur about between 1 01:01:58.000 --> 01:02:00.010 align:middle line:84% and 1,000 and 1,000 and 2,000 base pairs. 01:02:00.010 --> 01:02:03.520 align:middle line:84% They're approximately three-- maybe three to five million 01:02:03.520 --> 01:02:04.690 align:middle line:90% in the human genome. 01:02:04.690 --> 01:02:08.470 align:middle line:84% And it's using these as the primary source 01:02:08.470 --> 01:02:11.710 align:middle line:84% of genetic variation that we're actually going 01:02:11.710 --> 01:02:13.390 align:middle line:90% at trying to map these genes. 01:02:13.390 --> 01:02:15.310 align:middle line:84% There's a whole host of questions 01:02:15.310 --> 01:02:17.950 align:middle line:90% about how do you pick SNPs? 01:02:17.950 --> 01:02:20.200 align:middle line:84% We wrote a paper together, Zach and I, 01:02:20.200 --> 01:02:25.540 align:middle line:84% with some of our colleagues, about haplotype tagging SNPs. 01:02:25.540 --> 01:02:29.020 align:middle line:84% There's other approaches to using linkage disequilibrium 01:02:29.020 --> 01:02:31.310 align:middle line:84% to define the SNPs that you want to genotype. 01:02:31.310 --> 01:02:35.500 align:middle line:84% So lots of issues there, where bioinformatics is interfacing 01:02:35.500 --> 01:02:37.090 align:middle line:90% with human genetics. 01:02:37.090 --> 01:02:40.270 align:middle line:90% And no one really knows-- 01:02:40.270 --> 01:02:42.010 align:middle line:90% this is probably not 30 million. 01:02:42.010 --> 01:02:44.980 align:middle line:90% This is probably three. 01:02:44.980 --> 01:02:50.140 align:middle line:84% But no one really knows how many of these SNPs 01:02:50.140 --> 01:02:53.800 align:middle line:84% are actually coding, and I think everybody 01:02:53.800 --> 01:02:56.650 align:middle line:84% does know that there are more than coding 01:02:56.650 --> 01:02:58.300 align:middle line:90% SNPs that are important. 01:02:58.300 --> 01:03:01.450 align:middle line:84% Our motor SNPs are important, our coding SNPs are important, 01:03:01.450 --> 01:03:03.880 align:middle line:84% SNPs in the 3 prime UTR are important because they're 01:03:03.880 --> 01:03:07.150 align:middle line:84% going to change transcription factor binding 01:03:07.150 --> 01:03:09.160 align:middle line:84% and potentially change message level-- 01:03:09.160 --> 01:03:12.700 align:middle line:84% whole host of different-- and any one SNP 01:03:12.700 --> 01:03:14.740 align:middle line:84% is probably, in and of itself, isn't 01:03:14.740 --> 01:03:18.850 align:middle line:84% going to change function in a gene all that dramatically. 01:03:18.850 --> 01:03:24.010 align:middle line:84% So people are going towards this idea of analyzing data 01:03:24.010 --> 01:03:28.210 align:middle line:84% at the molecular level by looking at relevant functional 01:03:28.210 --> 01:03:28.900 align:middle line:90% haplotypes. 01:03:28.900 --> 01:03:34.510 align:middle line:84% If you've got a couple of SNPs in the promoter and another 01:03:34.510 --> 01:03:37.690 align:middle line:84% that's a nonsynonymous C SNP and an exon, 01:03:37.690 --> 01:03:46.240 align:middle line:84% another that's at a splice site, another that's in 3 01:03:46.240 --> 01:03:49.570 align:middle line:84% prime UTR that's determining message level of stability-- 01:03:49.570 --> 01:03:51.340 align:middle line:84% you combine all of those SNPs to try 01:03:51.340 --> 01:03:55.120 align:middle line:84% to get an effect across that whole gene in terms 01:03:55.120 --> 01:03:58.420 align:middle line:84% of looking at that gene and its impact on phenotype. 01:03:58.420 --> 01:04:01.130 align:middle line:90% 01:04:01.130 --> 01:04:03.370 align:middle line:84% So this is just a little bit about data analysis. 01:04:03.370 --> 01:04:09.760 align:middle line:84% You can either look at continuous quantitative traits 01:04:09.760 --> 01:04:10.960 align:middle line:90% or qualitative traits. 01:04:10.960 --> 01:04:13.990 align:middle line:84% There are parametric and nonparametric approaches 01:04:13.990 --> 01:04:15.940 align:middle line:90% to this. 01:04:15.940 --> 01:04:21.130 align:middle line:84% Then you use all this stuff you actually find the gene. 01:04:21.130 --> 01:04:22.810 align:middle line:84% I think the people are not doing-- 01:04:22.810 --> 01:04:25.150 align:middle line:84% the initial work was done with the [INAUDIBLE] clones, 01:04:25.150 --> 01:04:27.970 align:middle line:84% but now we're past the idea of doing that because there's 01:04:27.970 --> 01:04:30.610 align:middle line:84% enough markers with a HapMap project across the genome 01:04:30.610 --> 01:04:34.360 align:middle line:84% that we can go into almost any region now in the genome, 01:04:34.360 --> 01:04:37.930 align:middle line:84% and we can come up with validated SNPs 01:04:37.930 --> 01:04:40.480 align:middle line:84% across that region so that we can actually 01:04:40.480 --> 01:04:43.240 align:middle line:84% pick SNPs and genotype them and go directly-- 01:04:43.240 --> 01:04:48.327 align:middle line:84% and this is what's accelerating the pace of positional cloning 01:04:48.327 --> 01:04:48.910 align:middle line:90% at the moment. 01:04:48.910 --> 01:04:51.740 align:middle line:90% 01:04:51.740 --> 01:04:53.323 align:middle line:84% So these are some of the things that-- 01:04:53.323 --> 01:04:54.782 align:middle line:84% I haven't really talked about this. 01:04:54.782 --> 01:04:56.290 align:middle line:84% This is introductory, this lecture. 01:04:56.290 --> 01:04:58.510 align:middle line:84% But you really get into this-- how 01:04:58.510 --> 01:05:02.770 align:middle line:84% do you do haplotype analysis, ancestral haplotype analysis 01:05:02.770 --> 01:05:06.130 align:middle line:84% or linkage disequilibrium mapping, 01:05:06.130 --> 01:05:08.500 align:middle line:84% molecular methods or tissue expression-- all 01:05:08.500 --> 01:05:11.870 align:middle line:84% of these things can potentially be helpful in the fine mapping 01:05:11.870 --> 01:05:12.370 align:middle line:90% process. 01:05:12.370 --> 01:05:15.010 align:middle line:84% We've been very interested and have 01:05:15.010 --> 01:05:16.810 align:middle line:84% a project with Zach, where we wanted 01:05:16.810 --> 01:05:21.730 align:middle line:84% to use mouse expression and mouse QTL analysis 01:05:21.730 --> 01:05:24.227 align:middle line:84% to help us with human positional cloning. 01:05:24.227 --> 01:05:25.810 align:middle line:84% We're not sure if our project is going 01:05:25.810 --> 01:05:28.018 align:middle line:84% to be funded so we don't know if we're actually going 01:05:28.018 --> 01:05:29.200 align:middle line:90% to get a chance to do that. 01:05:29.200 --> 01:05:30.742 align:middle line:84% At the end of the day, you want to be 01:05:30.742 --> 01:05:33.400 align:middle line:84% able to look at the impact of polymorphic variation 01:05:33.400 --> 01:05:36.700 align:middle line:84% in the gene that you found and see whether that polymorphic-- 01:05:36.700 --> 01:05:39.430 align:middle line:84% how much of the phenotypic variance 01:05:39.430 --> 01:05:43.090 align:middle line:84% is explained by that polymorphism. 01:05:43.090 --> 01:05:45.170 align:middle line:84% And that gets back to this question of, 01:05:45.170 --> 01:05:48.190 align:middle line:84% well, you found a gene by positional cloning. 01:05:48.190 --> 01:05:49.990 align:middle line:84% How do it's really a significant gene? 01:05:49.990 --> 01:05:52.900 align:middle line:84% Well, does it replicate across different populations 01:05:52.900 --> 01:05:54.490 align:middle line:90% in different conditions? 01:05:54.490 --> 01:05:57.430 align:middle line:84% Is it important in different kinds of asthma? 01:05:57.430 --> 01:06:00.160 align:middle line:84% Does it seem to be explaining a significant amount 01:06:00.160 --> 01:06:01.280 align:middle line:90% of the variation? 01:06:01.280 --> 01:06:02.870 align:middle line:90% So this is one example. 01:06:02.870 --> 01:06:05.780 align:middle line:84% It's a poor example because it's not a really good one. 01:06:05.780 --> 01:06:09.280 align:middle line:84% This is a gene, CD-14, that we genotyped in the program 01:06:09.280 --> 01:06:10.870 align:middle line:90% in genomic applications. 01:06:10.870 --> 01:06:19.810 align:middle line:84% This gene is the gene that binds LPS, or lipopolysaccharide, 01:06:19.810 --> 01:06:23.050 align:middle line:84% to the membrane of the monocyte, and then transduces 01:06:23.050 --> 01:06:28.220 align:middle line:84% that signal to the T cell to produce Th1 cytokines. 01:06:28.220 --> 01:06:30.800 align:middle line:84% So we found a polymorphism in this gene 01:06:30.800 --> 01:06:33.470 align:middle line:84% as part of the program in genomic applications. 01:06:33.470 --> 01:06:35.210 align:middle line:90% It's a CDP polymorphism. 01:06:35.210 --> 01:06:39.410 align:middle line:90% So here's the T variant. 01:06:39.410 --> 01:06:41.390 align:middle line:90% Here's the heterozygote. 01:06:41.390 --> 01:06:42.770 align:middle line:90% And here's the C. 01:06:42.770 --> 01:06:46.790 align:middle line:84% And you can see that if you look at a dominant model, 01:06:46.790 --> 01:06:50.210 align:middle line:84% where the C's are together, anybody 01:06:50.210 --> 01:06:56.330 align:middle line:84% that has a C genotype actually is 01:06:56.330 --> 01:06:59.060 align:middle line:84% likely to have more positive skin tests than those who 01:06:59.060 --> 01:07:05.090 align:middle line:84% are T, T. And that genetic variation is associated 01:07:05.090 --> 01:07:10.530 align:middle line:84% with variation in soluble CD-14 levels in peripheral blood. 01:07:10.530 --> 01:07:12.470 align:middle line:84% So there's a relationship between 01:07:12.470 --> 01:07:20.930 align:middle line:84% genotype and intermediate phenotype, and a relationship 01:07:20.930 --> 01:07:25.820 align:middle line:90% to allergy, ultimately. 01:07:25.820 --> 01:07:27.740 align:middle line:84% AUDIENCE: I'm sorry, is that supposed to show 01:07:27.740 --> 01:07:29.750 align:middle line:90% a difference between the two? 01:07:29.750 --> 01:07:32.310 align:middle line:84% SCOTT WEISS: It's small, but it was significant. 01:07:32.310 --> 01:07:35.690 align:middle line:84% Well, I think the point here is that this is one SNP. 01:07:35.690 --> 01:07:38.480 align:middle line:84% This gets back to the point that it's not even 01:07:38.480 --> 01:07:42.230 align:middle line:84% a haplotype in this gene, and still there is-- 01:07:42.230 --> 01:07:43.880 align:middle line:90% and these are modest numbers. 01:07:43.880 --> 01:07:44.930 align:middle line:90% They're not huge. 01:07:44.930 --> 01:07:48.620 align:middle line:84% But there was clearly a difference, probably 01:07:48.620 --> 01:07:50.300 align:middle line:84% the level of difference you'd expect 01:07:50.300 --> 01:07:51.560 align:middle line:90% if it was just a single SNP. 01:07:51.560 --> 01:07:54.230 align:middle line:84% I mean, none of these effects are 01:07:54.230 --> 01:07:57.380 align:middle line:84% going to be very large at the level of an individual variant. 01:07:57.380 --> 01:08:00.020 align:middle line:84% At the level of a gene with a haplotype, 01:08:00.020 --> 01:08:03.560 align:middle line:84% with a really significant gene, maybe so, but certainly not 01:08:03.560 --> 01:08:04.790 align:middle line:90% one SNP. 01:08:04.790 --> 01:08:07.970 align:middle line:84% So these are some of the skills that if you guys want 01:08:07.970 --> 01:08:12.050 align:middle line:84% to do this work, if you were going to come to my laboratory, 01:08:12.050 --> 01:08:15.740 align:middle line:84% I would want you to know something about it. 01:08:15.740 --> 01:08:19.123 align:middle line:84% You'd want to be want to know something about this. 01:08:19.123 --> 01:08:20.540 align:middle line:84% And how to genotype and apply this 01:08:20.540 --> 01:08:24.140 align:middle line:84% to the disease, study design, statistical methodology, 01:08:24.140 --> 01:08:26.420 align:middle line:84% phenotyping, environmental exposures, and I 01:08:26.420 --> 01:08:29.450 align:middle line:84% probably ought to add to this list bioinformatics 01:08:29.450 --> 01:08:31.520 align:middle line:84% because without good bioinformatics skills, 01:08:31.520 --> 01:08:34.700 align:middle line:90% you're going to be lost. 01:08:34.700 --> 01:08:38.120 align:middle line:84% And it's hard to know exactly where on the spectrum 01:08:38.120 --> 01:08:39.560 align:middle line:90% people want to-- 01:08:39.560 --> 01:08:43.250 align:middle line:84% if could do this and never have anything 01:08:43.250 --> 01:08:46.970 align:middle line:84% to do with the phenotyping and just focus on the functional 01:08:46.970 --> 01:08:52.069 align:middle line:84% variation from genes that these guys are actually finding, 01:08:52.069 --> 01:08:56.300 align:middle line:84% or you might situate yourself somewhere in the middle, 01:08:56.300 --> 01:08:58.729 align:middle line:84% I've got people in my lab that are doing just this 01:08:58.729 --> 01:09:02.510 align:middle line:84% and very few people that are doing just this, 01:09:02.510 --> 01:09:05.580 align:middle line:84% but I have some that are sitting in the middle. 01:09:05.580 --> 01:09:07.560 align:middle line:84% So where is this going in the future? 01:09:07.560 --> 01:09:11.660 align:middle line:84% I mean, I think that what's driving the field 01:09:11.660 --> 01:09:14.390 align:middle line:84% is high-throughput sequencing and high-throughput 01:09:14.390 --> 01:09:16.970 align:middle line:84% genotyping combined with bioinformatics 01:09:16.970 --> 01:09:19.939 align:middle line:84% in the presence of having lots of populations 01:09:19.939 --> 01:09:21.109 align:middle line:90% to do this kind of work. 01:09:21.109 --> 01:09:23.090 align:middle line:84% That's what's really necessary is 01:09:23.090 --> 01:09:25.130 align:middle line:84% you've got to have well phenotype populations. 01:09:25.130 --> 01:09:27.710 align:middle line:84% In my lab, these are all the different populations 01:09:27.710 --> 01:09:28.970 align:middle line:90% that we have for asthma. 01:09:28.970 --> 01:09:30.800 align:middle line:84% We've got these extended pedigrees. 01:09:30.800 --> 01:09:32.210 align:middle line:90% We've got affected sib pairs. 01:09:32.210 --> 01:09:37.220 align:middle line:84% We've got trios, and we've got individual cases and controls 01:09:37.220 --> 01:09:40.850 align:middle line:84% so that we can test the genes in multiple different populations 01:09:40.850 --> 01:09:45.073 align:middle line:90% and under different conditions. 01:09:45.073 --> 01:09:46.490 align:middle line:84% So why don't I stop there, and I'd 01:09:46.490 --> 01:09:49.130 align:middle line:84% be glad to answer any questions that people have about 01:09:49.130 --> 01:09:52.577 align:middle line:90% any of the things that I said. 01:09:52.577 --> 01:09:55.881 align:middle line:90% AUDIENCE: [INAUDIBLE] 01:09:55.881 --> 01:09:58.757 align:middle line:90% 01:09:58.757 --> 01:10:00.215 align:middle line:84% SCOTT WEISS: Structure of proteins? 01:10:00.215 --> 01:10:02.750 align:middle line:90% AUDIENCE: [INAUDIBLE] 01:10:02.750 --> 01:10:05.390 align:middle line:84% SCOTT WEISS: Well, I mean I think that means is that you 01:10:05.390 --> 01:10:08.660 align:middle line:90% actually getting into-- 01:10:08.660 --> 01:10:14.780 align:middle line:84% once you've got a relationship with a gene, what 01:10:14.780 --> 01:10:23.660 align:middle line:84% you have to do is really get down and figure out 01:10:23.660 --> 01:10:26.360 align:middle line:84% what are the variants in that gene and what are they doing, 01:10:26.360 --> 01:10:29.150 align:middle line:90% and that can proceed. 01:10:29.150 --> 01:10:31.220 align:middle line:84% Human genetics can contribute to that 01:10:31.220 --> 01:10:33.060 align:middle line:84% at the level of genetic association. 01:10:33.060 --> 01:10:36.200 align:middle line:84% So, for example, Lori Glimcher, who's 01:10:36.200 --> 01:10:39.200 align:middle line:84% an immunologist at the School of Public Health, 01:10:39.200 --> 01:10:46.250 align:middle line:84% identified a gene that controls T-cell differentiation. 01:10:46.250 --> 01:10:49.610 align:middle line:84% It's TBX-21, or T-bet is the name of the gene. 01:10:49.610 --> 01:10:51.830 align:middle line:84% And she created a knockout mouse, 01:10:51.830 --> 01:10:55.820 align:middle line:84% and when you knock this gene out in the mouse, 01:10:55.820 --> 01:10:58.070 align:middle line:84% you get tremendous airways responsiveness 01:10:58.070 --> 01:11:00.470 align:middle line:90% and allergic inflammation. 01:11:00.470 --> 01:11:03.460 align:middle line:84% It looks like an asthma gene in the mouse. 01:11:03.460 --> 01:11:08.390 align:middle line:84% And so we sequenced that gene, and then we 01:11:08.390 --> 01:11:15.110 align:middle line:84% started to look at-- we found a variant in the gene that's 01:11:15.110 --> 01:11:16.370 align:middle line:90% in the coding region. 01:11:16.370 --> 01:11:19.850 align:middle line:84% It's a nonsynonymous C SNP in the coding region. 01:11:19.850 --> 01:11:20.760 align:middle line:90% It's very rare. 01:11:20.760 --> 01:11:22.800 align:middle line:84% It's only occurs in about 3% of people. 01:11:22.800 --> 01:11:29.900 align:middle line:84% But it turns out that coding region variant 01:11:29.900 --> 01:11:35.750 align:middle line:84% determines which patients who get inhaled steroids 01:11:35.750 --> 01:11:37.010 align:middle line:90% get better. 01:11:37.010 --> 01:11:40.130 align:middle line:84% The people that have that variant 01:11:40.130 --> 01:11:43.100 align:middle line:84% and get inhaled steroids have their airways responsiveness 01:11:43.100 --> 01:11:45.253 align:middle line:90% completely return to normal. 01:11:45.253 --> 01:11:46.670 align:middle line:84% AUDIENCE: Has this been published? 01:11:46.670 --> 01:11:48.560 align:middle line:84% SCOTT WEISS: We're about to submit it to The Lancet. 01:11:48.560 --> 01:11:50.240 align:middle line:84% We're actually working with Lori to-- 01:11:50.240 --> 01:11:51.657 align:middle line:84% it's pretty exciting [INAUDIBLE].. 01:11:51.657 --> 01:11:54.355 align:middle line:84% And it's exciting because it's an example of how 01:11:54.355 --> 01:11:55.730 align:middle line:84% you can actually-- you don't have 01:11:55.730 --> 01:11:57.690 align:middle line:90% to even go to the animal model. 01:11:57.690 --> 01:12:02.600 align:middle line:84% And so she, then, has created her mouse model. 01:12:02.600 --> 01:12:05.420 align:middle line:84% She started to do some experiments with steroids, 01:12:05.420 --> 01:12:08.780 align:middle line:84% and steroids are probably important in controlling 01:12:08.780 --> 01:12:10.400 align:middle line:90% T-bet expression. 01:12:10.400 --> 01:12:12.180 align:middle line:90% And she didn't know that. 01:12:12.180 --> 01:12:15.380 align:middle line:84% So that's an example of structure-function 01:12:15.380 --> 01:12:18.260 align:middle line:84% relationships, where you're trying to figure out 01:12:18.260 --> 01:12:19.830 align:middle line:90% what a gene actually does. 01:12:19.830 --> 01:12:21.830 align:middle line:84% And it is important to recognize, 01:12:21.830 --> 01:12:26.630 align:middle line:84% there are some genes that have been around for a while, 01:12:26.630 --> 01:12:30.860 align:middle line:84% and people still don't know there's 01:12:30.860 --> 01:12:32.960 align:middle line:84% a relationship to a disease phenotype, 01:12:32.960 --> 01:12:35.240 align:middle line:90% but we don't know how they work. 01:12:35.240 --> 01:12:37.670 align:middle line:84% So figuring out that structure-function stuff 01:12:37.670 --> 01:12:40.010 align:middle line:84% can take a long time, potentially. 01:12:40.010 --> 01:12:41.660 align:middle line:84% And doing the genetic association 01:12:41.660 --> 01:12:43.670 align:middle line:84% and the fine mapping may actually now 01:12:43.670 --> 01:12:47.930 align:middle line:84% proceed at a faster pace and not take as much time. 01:12:47.930 --> 01:12:56.480 align:middle line:84% But I think you can actually do a lot of structure function 01:12:56.480 --> 01:12:57.590 align:middle line:90% stuff. 01:12:57.590 --> 01:13:00.500 align:middle line:84% Usually, what we do is when we get an association, 01:13:00.500 --> 01:13:03.140 align:middle line:90% we will type every damn-- 01:13:03.140 --> 01:13:05.870 align:middle line:84% we'll sequence that gene, we'll type every damn variant 01:13:05.870 --> 01:13:08.450 align:middle line:84% we can find in that gene in population 01:13:08.450 --> 01:13:10.640 align:middle line:84% and look at everything that could be related 01:13:10.640 --> 01:13:13.730 align:middle line:84% to an interesting phenotype because we're searching 01:13:13.730 --> 01:13:17.000 align:middle line:84% for clues to how to help our molecular biology 01:13:17.000 --> 01:13:21.740 align:middle line:84% colleagues in trying to help them figure out what 01:13:21.740 --> 01:13:22.955 align:middle line:90% the gene is actually doing. 01:13:22.955 --> 01:13:25.850 align:middle line:90% 01:13:25.850 --> 01:13:29.660 align:middle line:84% AUDIENCE: [INAUDIBLE] of that group [INAUDIBLE] 01:13:29.660 --> 01:13:32.768 align:middle line:84% see how [INAUDIBLE] changes [INAUDIBLE].. 01:13:32.768 --> 01:13:34.310 align:middle line:84% SCOTT WEISS: Well, you could do that. 01:13:34.310 --> 01:13:37.040 align:middle line:84% I mean, we're trying to work with this guy. 01:13:37.040 --> 01:13:39.770 align:middle line:84% He's got people in his lab who have 01:13:39.770 --> 01:13:44.810 align:middle line:90% ideas about how to get clues. 01:13:44.810 --> 01:13:47.720 align:middle line:84% So like the stuff that [INAUDIBLE] showed you last 01:13:47.720 --> 01:13:49.760 align:middle line:90% week, that gene CRHR1-- 01:13:49.760 --> 01:13:53.060 align:middle line:90% 01:13:53.060 --> 01:13:55.820 align:middle line:84% we know a relationship to steroid treatment response, 01:13:55.820 --> 01:13:58.550 align:middle line:84% but we don't know what the variant is in the gene. 01:13:58.550 --> 01:14:00.350 align:middle line:84% So we sequenced the gene completely, 01:14:00.350 --> 01:14:04.640 align:middle line:84% and now we've got two INDELs in that gene 01:14:04.640 --> 01:14:07.670 align:middle line:84% that sit right at intron/exon junctions. 01:14:07.670 --> 01:14:10.970 align:middle line:84% So the presumed-- what we're thinking 01:14:10.970 --> 01:14:16.610 align:middle line:84% is that those insertion deletion polymorphisms may be changing 01:14:16.610 --> 01:14:18.197 align:middle line:90% alternative splice sites. 01:14:18.197 --> 01:14:20.030 align:middle line:84% So we're going to have to try to prove that. 01:14:20.030 --> 01:14:22.070 align:middle line:84% That's one of the hypotheses that we're 01:14:22.070 --> 01:14:24.080 align:middle line:84% going to investigate in the renewal of the grant 01:14:24.080 --> 01:14:25.340 align:middle line:90% is trying to look at that. 01:14:25.340 --> 01:14:28.340 align:middle line:90% 01:14:28.340 --> 01:14:36.050 align:middle line:84% So you have to let the gene tell you where its variation is 01:14:36.050 --> 01:14:42.630 align:middle line:84% and how it might be contributing to phenotype. 01:14:42.630 --> 01:14:45.500 align:middle line:84% And so the first thing, usually, is 01:14:45.500 --> 01:14:47.180 align:middle line:90% to sequence the gene completely. 01:14:47.180 --> 01:14:49.880 align:middle line:84% Second thing would be to then do a very careful analysis 01:14:49.880 --> 01:14:53.390 align:middle line:84% of the new variants and the resequenced variants 01:14:53.390 --> 01:14:57.583 align:middle line:84% that you found in relationship to the phenotype of interest 01:14:57.583 --> 01:14:59.000 align:middle line:84% or phenotypes of interest, and see 01:14:59.000 --> 01:15:04.520 align:middle line:84% if you can find either haplotypes or individual SNPs 01:15:04.520 --> 01:15:07.100 align:middle line:84% or insertion-deletion polymorphisms 01:15:07.100 --> 01:15:09.830 align:middle line:84% or transcription factor binding sites or things 01:15:09.830 --> 01:15:13.700 align:middle line:84% that could potentially explain the genetic association. 01:15:13.700 --> 01:15:15.290 align:middle line:84% So then you can do that, and then you 01:15:15.290 --> 01:15:19.160 align:middle line:84% have to go into an animal model and test those in a more 01:15:19.160 --> 01:15:20.953 align:middle line:90% rigorous way, usually. 01:15:20.953 --> 01:15:22.370 align:middle line:84% AUDIENCE: How far off do you think 01:15:22.370 --> 01:15:26.480 align:middle line:84% is the day when a commission [INAUDIBLE] 01:15:26.480 --> 01:15:30.290 align:middle line:84% will be able to come to you and say, I have a disease. 01:15:30.290 --> 01:15:35.130 align:middle line:84% I have 500 [INAUDIBLE] cases and 500 controls. 01:15:35.130 --> 01:15:43.533 align:middle line:84% I think [INAUDIBLE] on the long arm of a chromosome. 01:15:43.533 --> 01:15:45.360 align:middle line:90% And I want to [INAUDIBLE]. 01:15:45.360 --> 01:15:50.550 align:middle line:90% 01:15:50.550 --> 01:15:52.913 align:middle line:84% SCOTT WEISS: So you're really asking, the question is-- 01:15:52.913 --> 01:15:54.330 align:middle line:84% I think the question you're asking 01:15:54.330 --> 01:15:59.195 align:middle line:84% is how far away is whole genome association? 01:15:59.195 --> 01:16:01.320 align:middle line:84% AUDIENCE: How far is whole genome association where 01:16:01.320 --> 01:16:04.440 align:middle line:84% it's within the reach of significant but not impossible 01:16:04.440 --> 01:16:07.410 align:middle line:90% clinical studies. 01:16:07.410 --> 01:16:10.110 align:middle line:90% SCOTT WEISS: Max three years. 01:16:10.110 --> 01:16:12.626 align:middle line:90% Max. 01:16:12.626 --> 01:16:13.990 align:middle line:90% AUDIENCE: All right. 01:16:13.990 --> 01:16:17.470 align:middle line:84% SCOTT WEISS: I mean, George Church-- 01:16:17.470 --> 01:16:20.080 align:middle line:84% it's all about the genotyping costs, Zach. 01:16:20.080 --> 01:16:22.120 align:middle line:90% AUDIENCE: [INAUDIBLE] 01:16:22.120 --> 01:16:23.650 align:middle line:90% SCOTT WEISS: I mean, he-- 01:16:23.650 --> 01:16:31.970 align:middle line:84% listen, he thinks he's close to the $1,000 genome. 01:16:31.970 --> 01:16:34.720 align:middle line:84% So if he's really close to the $1,000 genome 01:16:34.720 --> 01:16:39.190 align:middle line:84% and SNP genotyping costs really drop, continue to drop 01:16:39.190 --> 01:16:41.440 align:middle line:84% as dramatically as they've dropped over the last three 01:16:41.440 --> 01:16:46.240 align:middle line:84% years, I would see whole genome association 01:16:46.240 --> 01:16:54.670 align:middle line:84% being within the range of a reasonable budget in a two 01:16:54.670 --> 01:16:55.988 align:middle line:90% or three-year period of time. 01:16:55.988 --> 01:16:58.280 align:middle line:84% AUDIENCE: All right, on that note, thank you very much, 01:16:58.280 --> 01:17:00.410 align:middle line:90% Scott. 01:17:00.410 --> 01:17:01.960 align:middle line:90% And--