1 00:00:06,950 --> 00:00:07,750 ERIC LANDER: Good morning. 2 00:00:07,750 --> 00:00:09,000 Good morning. 3 00:00:17,480 --> 00:00:27,330 So last time, we ran into a problem. 4 00:00:27,330 --> 00:00:31,760 We had Mendel, my hero Mendel, this MIT-like mathematical, 5 00:00:31,760 --> 00:00:40,200 physical monk, had developed this gorgeous theory of 6 00:00:40,200 --> 00:00:41,500 particles of inheritance. 7 00:00:41,500 --> 00:00:43,330 He didn't use the word "gene" yet. 8 00:00:43,330 --> 00:00:45,295 Gene doesn't get invented for much longer. 9 00:00:48,140 --> 00:00:53,760 For every trait, you had two such particles. 10 00:00:53,760 --> 00:00:53,922 You gave one to your offspring. 11 00:00:53,922 --> 00:00:56,300 Each of the parents gave one to their offspring. 12 00:00:56,300 --> 00:01:00,000 And that's how each offspring gets two of them. 13 00:01:00,000 --> 00:01:05,239 That choice of which of the two alleles to transmit to 14 00:01:05,239 --> 00:01:07,790 your offspring is a random draw. 15 00:01:07,790 --> 00:01:11,470 And that explains beautifully, for example, the 3-to-1 16 00:01:11,470 --> 00:01:14,130 segregation pattern that Mendel saw-- 17 00:01:14,130 --> 00:01:16,590 gorgeous. 18 00:01:16,590 --> 00:01:20,620 We put that model, which was an ex post facto model, a 19 00:01:20,620 --> 00:01:24,200 model made after the data were available, to a test. 20 00:01:24,200 --> 00:01:27,100 You guys insisted we had to test it before you would 21 00:01:27,100 --> 00:01:28,280 publish it. 22 00:01:28,280 --> 00:01:31,910 And it holds up pretty well, making pretty surprising 23 00:01:31,910 --> 00:01:34,000 predictions that you would otherwise 24 00:01:34,000 --> 00:01:35,730 not have ever expected. 25 00:01:35,730 --> 00:01:39,660 Like amongst that 3 to 1, the threes are not all the same. 26 00:01:39,660 --> 00:01:45,430 Some of those round peas were homozygous 27 00:01:45,430 --> 00:01:48,060 for the big-R allele. 28 00:01:48,060 --> 00:01:51,910 And in selfing, they'll never produce wrinkled peas. 29 00:01:51,910 --> 00:01:56,310 But 2/3 of those round peas were heterozygous. 30 00:01:56,310 --> 00:02:00,070 And when you self them, you get 1/4 wrinkled peas. 31 00:02:00,070 --> 00:02:04,170 So the wacky prediction that 1/3 of the rounds will give 32 00:02:04,170 --> 00:02:07,050 rise to no wrinkleds, and 2/3 of the rounds will give rise 33 00:02:07,050 --> 00:02:11,050 to 1/4 wrinkleds is a surprising prediction and, 34 00:02:11,050 --> 00:02:14,570 therefore, something that bears stating as a scientific 35 00:02:14,570 --> 00:02:16,491 prediction. 36 00:02:16,491 --> 00:02:20,820 And so all those kind of predictions can emerge. 37 00:02:20,820 --> 00:02:26,040 Then we turned to the question of two factors. 38 00:02:26,040 --> 00:02:29,180 I'm practicing using our words here-- 39 00:02:29,180 --> 00:02:32,800 homozygous, heterozygous, alleles, et cetera. 40 00:02:32,800 --> 00:02:37,130 We now turn to two traits, two phenotypes. 41 00:02:37,130 --> 00:02:41,550 We have two phenotypes segregating. 42 00:02:41,550 --> 00:02:45,270 We had round and wrinkled, and we had green and yellow. 43 00:02:45,270 --> 00:02:49,390 And Mendel determined, rather brilliantly, that they 44 00:02:49,390 --> 00:02:53,890 segregated, were transmitted independently of each other. 45 00:02:53,890 --> 00:02:56,850 There was no correlation between which alleles you got 46 00:02:56,850 --> 00:02:59,590 at round and which alleles you got at wrinkled. 47 00:02:59,590 --> 00:03:02,450 That was pretty cool. 48 00:03:02,450 --> 00:03:05,480 Let's just go over that, because there's a real tension 49 00:03:05,480 --> 00:03:23,330 to be resolved because we have Mendel's second law versus the 50 00:03:23,330 --> 00:03:24,580 chromosome theory. 51 00:03:32,500 --> 00:03:34,260 So Mendel's second law-- 52 00:03:34,260 --> 00:03:40,005 let's just go back to it-- in the F0 generation, we had our 53 00:03:40,005 --> 00:03:50,030 round green peas, genotype big R, big R, big G, big G. We had 54 00:03:50,030 --> 00:03:59,370 our wrinkled yellow peas, genotype little r, little r, 55 00:03:59,370 --> 00:04:01,430 little g, little g. 56 00:04:01,430 --> 00:04:08,420 We cross them together, we get F1 double heterozygotes, big 57 00:04:08,420 --> 00:04:11,120 R, little r, big G, little g. 58 00:04:11,120 --> 00:04:18,850 We then perform a back cross or test cross to the doubly 59 00:04:18,850 --> 00:04:23,170 homozygous parent with the two recessive phenotypes-- 60 00:04:23,170 --> 00:04:25,280 we're practicing our words here. 61 00:04:25,280 --> 00:04:26,680 And what do we get? 62 00:04:26,680 --> 00:04:29,030 Well, we get certain options. 63 00:04:29,030 --> 00:04:33,290 As we said, the gametes that emerge from this parent on the 64 00:04:33,290 --> 00:04:37,205 left could be of the following types. 65 00:04:41,420 --> 00:04:44,590 The gametes from the parent on the right 66 00:04:44,590 --> 00:04:47,120 are all those alleles-- 67 00:04:49,980 --> 00:04:51,230 Are those the recessive alleles? 68 00:04:54,140 --> 00:04:54,610 No. 69 00:04:54,610 --> 00:04:57,140 You'll forget this over time, but just to make me 70 00:04:57,140 --> 00:04:58,000 comfortable-- 71 00:04:58,000 --> 00:05:00,460 they're actually the alleles associated with the recessive 72 00:05:00,460 --> 00:05:02,280 phenotype that we're talking about. 73 00:05:02,280 --> 00:05:04,830 Because you know-- but will forget, I assure you, because 74 00:05:04,830 --> 00:05:06,370 all my colleagues in the biology department have 75 00:05:06,370 --> 00:05:07,210 forgotten-- 76 00:05:07,210 --> 00:05:09,630 that they could also control multiple other phenotypes, 77 00:05:09,630 --> 00:05:11,760 some of which could be dominant. 78 00:05:11,760 --> 00:05:12,450 But that's OK. 79 00:05:12,450 --> 00:05:14,070 I forgive you in advance that you'll call 80 00:05:14,070 --> 00:05:15,590 them recessive alleles. 81 00:05:15,590 --> 00:05:18,520 Anyway, you get this. 82 00:05:18,520 --> 00:05:23,673 And then these should occur at equal frequency of one to one 83 00:05:23,673 --> 00:05:24,923 to one to one. 84 00:05:28,860 --> 00:05:29,871 All right. 85 00:05:29,871 --> 00:05:36,420 Now, we had the chromosome theory. 86 00:05:36,420 --> 00:05:39,890 The chromosome theory, the observation of the 87 00:05:39,890 --> 00:05:43,260 choreography of chromosomes during meiosis-- 88 00:05:43,260 --> 00:05:46,730 chromosomes in meiosis-- 89 00:05:46,730 --> 00:05:47,980 looks like this. 90 00:05:50,790 --> 00:05:56,010 We have chromosomes lining up in pairs. 91 00:05:56,010 --> 00:05:57,490 Now, I'm not drawing it terribly well. 92 00:05:57,490 --> 00:06:04,220 But the two members of each pair are of the same size. 93 00:06:04,220 --> 00:06:06,890 But this pair could be bigger, and this 94 00:06:06,890 --> 00:06:07,900 pair could be smaller. 95 00:06:07,900 --> 00:06:09,920 It looks like they're really finding each other. 96 00:06:09,920 --> 00:06:12,620 These chromosomes are visibly different in shape. 97 00:06:12,620 --> 00:06:15,450 And so maybe I'll make this guy a little longer, just to 98 00:06:15,450 --> 00:06:18,020 indicate that, that it actually knows. 99 00:06:18,020 --> 00:06:24,390 Now an explanation, we said, for how Mendel's second law of 100 00:06:24,390 --> 00:06:27,330 independent assortment of two different phenotypes could 101 00:06:27,330 --> 00:06:29,665 occur is that, for example - round. 102 00:06:40,910 --> 00:06:44,130 It could be that the gene for roundness is located on 103 00:06:44,130 --> 00:06:46,030 chromosome number one. 104 00:06:46,030 --> 00:06:48,460 These are two copies of chromosome one that have been 105 00:06:48,460 --> 00:06:52,050 duplicated, each of which has the big-R allele. 106 00:06:52,050 --> 00:06:55,930 Next to it, two copies of chromosome one that have been 107 00:06:55,930 --> 00:07:00,380 duplicated, each of them having the little-r allele. 108 00:07:00,380 --> 00:07:02,920 Over here on chromosome number two. 109 00:07:02,920 --> 00:07:07,760 let's say, lies the gene for green or yellow. 110 00:07:07,760 --> 00:07:17,270 And in this picture here, the big G's are on the two copies 111 00:07:17,270 --> 00:07:19,210 of this chromosome number two that are here. 112 00:07:19,210 --> 00:07:21,770 The little g's are on the two copies of chromosome number 113 00:07:21,770 --> 00:07:24,150 two that are here. 114 00:07:24,150 --> 00:07:34,300 When the cell undergoes meiosis one, we get to the 115 00:07:34,300 --> 00:07:48,480 situation where we have big G, big G; little g, little g. 116 00:07:48,480 --> 00:07:54,650 We've got big R, big R; little r, little r. 117 00:07:54,650 --> 00:07:56,520 Could it have been the case that the big G was on the 118 00:07:56,520 --> 00:07:59,566 right and the little g was on the left? 119 00:07:59,566 --> 00:08:00,430 Yeah, of course. 120 00:08:00,430 --> 00:08:01,620 It's totally independent which way. 121 00:08:01,620 --> 00:08:03,250 I happened to draw it this way. 122 00:08:03,250 --> 00:08:06,010 But with probability 50%, it's the other way. 123 00:08:06,010 --> 00:08:07,860 That's why they're independent of each other. 124 00:08:07,860 --> 00:08:11,950 And then when it undergoes meiosis two that looks just 125 00:08:11,950 --> 00:08:19,100 like mitosis, we end up with our four gametes here. 126 00:08:21,940 --> 00:08:41,390 Sorry, our four gametes with big G, big G; little g, little 127 00:08:41,390 --> 00:08:49,330 g; big R, big R; little r, little r. 128 00:08:49,330 --> 00:08:54,950 And that accounts for the big G, big G; big R, big G; little 129 00:08:54,950 --> 00:08:56,690 r, little g gametes. 130 00:08:56,690 --> 00:08:59,600 And then when they went the other way, the organism would 131 00:08:59,600 --> 00:09:02,510 make a set of gametes that had the other combination of big 132 00:09:02,510 --> 00:09:04,610 R's with little g's. 133 00:09:04,610 --> 00:09:06,530 So that's perfectly fine. 134 00:09:06,530 --> 00:09:09,750 And because the second chromosome is independently 135 00:09:09,750 --> 00:09:12,030 ordered compared to the first chromosome-- when they line up 136 00:09:12,030 --> 00:09:13,220 at the midline, they don't really care 137 00:09:13,220 --> 00:09:14,390 which way they are-- 138 00:09:14,390 --> 00:09:16,960 that'll account for one to one to one to one. 139 00:09:16,960 --> 00:09:18,210 It's so straightforward. 140 00:09:22,920 --> 00:09:40,190 But what happens if, instead, both the roundness gene and 141 00:09:40,190 --> 00:09:43,425 the greenness gene live on the same chromosome? 142 00:09:46,850 --> 00:09:49,390 We'll have little r, little r there. 143 00:09:49,390 --> 00:09:56,060 We'll have big G, big G here; little g, little g. 144 00:09:56,060 --> 00:10:02,150 And then when they split, we'll end up with-- 145 00:10:02,150 --> 00:10:07,510 now, chromosome two has nothing that we 146 00:10:07,510 --> 00:10:08,500 care about on it. 147 00:10:08,500 --> 00:10:09,395 It has a lot of genes. 148 00:10:09,395 --> 00:10:12,050 In fact, there could be chromosomes three, four, five. 149 00:10:12,050 --> 00:10:13,530 I'm just not drawing them. 150 00:10:13,530 --> 00:10:15,290 But we're really going to focus on 151 00:10:15,290 --> 00:10:17,450 chromosome number one here. 152 00:10:17,450 --> 00:10:26,520 And you'll notice that here on chromosome one, the big G's or 153 00:10:26,520 --> 00:10:34,760 the little g's are coupled, physically coupled to each 154 00:10:34,760 --> 00:10:36,150 other, the bigs with the bigs and the 155 00:10:36,150 --> 00:10:37,820 smalls with the smalls. 156 00:10:37,820 --> 00:10:46,490 So now the kind of gametes that can emerge from this will 157 00:10:46,490 --> 00:10:50,410 only be big R, big G type or-- 158 00:10:50,410 --> 00:10:51,800 big G type-- 159 00:10:51,800 --> 00:10:56,270 or they'll be little r, little g type. 160 00:10:56,270 --> 00:10:58,980 We can't get the reverse combination. 161 00:10:58,980 --> 00:11:01,220 We can't get bigs and littles. 162 00:11:01,220 --> 00:11:10,040 So this, because these are independent of each other, 163 00:11:10,040 --> 00:11:19,840 will get us one to one to one to one. 164 00:11:19,840 --> 00:11:27,170 These are the big, little, and then these other 165 00:11:27,170 --> 00:11:29,610 combinations like that. 166 00:11:29,610 --> 00:11:39,070 This will get us only, if we look at the big R, big G; 167 00:11:39,070 --> 00:11:43,230 little r, little g; big R, little g; little r, big G; 168 00:11:43,230 --> 00:11:49,730 will get us one to one to zero to zero. 169 00:11:49,730 --> 00:11:52,490 Let's give a name to this type, the big R's and the 170 00:11:52,490 --> 00:11:53,920 little g's. 171 00:11:53,920 --> 00:11:58,200 Let's call that a recombinant type, OK? 172 00:11:58,200 --> 00:11:59,760 I'm just going to use that word for the moment. 173 00:11:59,760 --> 00:12:01,570 The recombinant types-- the bigs and the littles, and 174 00:12:01,570 --> 00:12:03,690 littles and the bigs-- 175 00:12:03,690 --> 00:12:05,350 we're not going to see any of them. 176 00:12:05,350 --> 00:12:08,710 This is a very strong difference between Mendel's 177 00:12:08,710 --> 00:12:10,170 second law and the chromosome theory. 178 00:12:10,170 --> 00:12:13,030 If the chromosome theory is right, and these chromosomes 179 00:12:13,030 --> 00:12:19,240 are physical entities that have integrity, they can't 180 00:12:19,240 --> 00:12:22,240 both be right. 181 00:12:22,240 --> 00:12:24,660 So that's a great thing in science, when you have two 182 00:12:24,660 --> 00:12:28,050 different models and they can't both be right, because 183 00:12:28,050 --> 00:12:29,210 you learn things then. 184 00:12:29,210 --> 00:12:30,740 You can test them. 185 00:12:30,740 --> 00:12:35,930 Now, Mendel tested this without actually knowing the 186 00:12:35,930 --> 00:12:37,400 chromosome theory. 187 00:12:37,400 --> 00:12:39,490 And he always got one to one to one to one for the seven 188 00:12:39,490 --> 00:12:40,740 traits he looked at. 189 00:12:43,620 --> 00:12:45,720 Was he just lucky that they happened to lie on the seven 190 00:12:45,720 --> 00:12:46,970 different chromosomes? 191 00:12:50,240 --> 00:12:52,040 Or is there some problem with this 192 00:12:52,040 --> 00:12:55,340 chromosome theory, or what? 193 00:12:55,340 --> 00:12:56,235 It took a while. 194 00:12:56,235 --> 00:12:59,220 And then, of course, everybody forgot about Mendel from 1865 195 00:12:59,220 --> 00:13:00,920 until the year 1900. 196 00:13:00,920 --> 00:13:04,210 In the year 1900, people begin to rediscover Mendel. 197 00:13:04,210 --> 00:13:06,010 Cytology has come along. 198 00:13:06,010 --> 00:13:10,750 In January of the year 1900, plant breeders start 199 00:13:10,750 --> 00:13:12,000 rediscovering Mendel. 200 00:13:16,780 --> 00:13:17,210 Sorry. 201 00:13:17,210 --> 00:13:18,562 Thank you. 202 00:13:18,562 --> 00:13:20,250 I see already. 203 00:13:20,250 --> 00:13:22,670 Thank you. 204 00:13:22,670 --> 00:13:24,910 I could tell by the look on your face that something must 205 00:13:24,910 --> 00:13:25,480 be wrong there. 206 00:13:25,480 --> 00:13:27,030 Good. 207 00:13:27,030 --> 00:13:30,860 Plant breeders start rediscovering Mendel. 208 00:13:30,860 --> 00:13:37,230 And in January of 1900, three different groups say, you 209 00:13:37,230 --> 00:13:38,710 know, we found these laws. 210 00:13:38,710 --> 00:13:40,230 And they're just like Mendel's laws-- 211 00:13:40,230 --> 00:13:42,680 which now everybody starts paying attention to. 212 00:13:42,680 --> 00:13:45,460 But plant breeding-- and people tried to do mice, and 213 00:13:45,460 --> 00:13:48,020 people tried to do rats. 214 00:13:48,020 --> 00:13:50,770 What turned out to be the winner, the place to really 215 00:13:50,770 --> 00:13:54,980 study genetics, was the fruit fly. 216 00:13:54,980 --> 00:13:59,490 Thomas Hunt Morgan at Columbia University decided, after he 217 00:13:59,490 --> 00:14:02,610 was really frustrated wasting years breeding mice and rats 218 00:14:02,610 --> 00:14:07,930 that just took too long, around, oh, I don't know, 1906 219 00:14:07,930 --> 00:14:12,160 or something like that, began to start breeding fruit flies. 220 00:14:12,160 --> 00:14:15,020 Fruit flies are the teeny little flies that when you 221 00:14:15,020 --> 00:14:18,660 open a banana or fruits or other kinds of 222 00:14:18,660 --> 00:14:20,220 things, you'll see them. 223 00:14:20,220 --> 00:14:26,250 And studying Drosophila gave us the answer to this question 224 00:14:26,250 --> 00:14:29,090 of what the problem is, how can it be that either it's 225 00:14:29,090 --> 00:14:31,930 independent or totally dependent? 226 00:14:31,930 --> 00:14:36,090 So we're going to talk about Drosophila melanogaster, the 227 00:14:36,090 --> 00:14:40,000 fruit fly, and the discovery of recombination. 228 00:14:47,350 --> 00:14:49,670 So now, Morgan-- 229 00:14:54,240 --> 00:14:56,920 I'm going to now start using fruit fly notation to give you 230 00:14:56,920 --> 00:15:00,180 some practice with fruit fly notation. 231 00:15:00,180 --> 00:15:02,380 We're not going to use big G's and little g's. 232 00:15:02,380 --> 00:15:04,690 They like to refer to the normal allele 233 00:15:04,690 --> 00:15:06,120 and the mutant allele. 234 00:15:06,120 --> 00:15:08,980 The normally allele is plus. 235 00:15:08,980 --> 00:15:11,930 The mutant allele gets some kind of a name. 236 00:15:11,930 --> 00:15:19,940 And so he had a female fly that was normal and normal at 237 00:15:19,940 --> 00:15:22,880 two different loci, two different genes. 238 00:15:22,880 --> 00:15:24,330 I haven't told you what the genes are. 239 00:15:24,330 --> 00:15:30,240 And the male fly had a black-colored body-- black 240 00:15:30,240 --> 00:15:32,230 across its whole body-- 241 00:15:32,230 --> 00:15:36,700 and its wings were shriveled and, 242 00:15:36,700 --> 00:15:38,840 therefore, called vestigial. 243 00:15:38,840 --> 00:15:45,570 So the phenotype here is black and vestigial, black body and 244 00:15:45,570 --> 00:15:47,400 vestigial wings. 245 00:15:47,400 --> 00:15:50,620 And this wasn't the normal appearance of a fly. 246 00:15:50,620 --> 00:15:54,330 So he took these females by these males, and he crossed 247 00:15:54,330 --> 00:15:55,270 them together. 248 00:15:55,270 --> 00:15:58,730 And he got an F1. 249 00:15:58,730 --> 00:16:05,600 And the F1 was black over vestigial, plus over plus. 250 00:16:05,600 --> 00:16:09,050 And what was their phenotype? 251 00:16:09,050 --> 00:16:12,720 Were they normal appearance, which is kind of a 252 00:16:12,720 --> 00:16:15,090 sandy-colored body and normal wings? 253 00:16:15,090 --> 00:16:19,820 Or were they this all black body and vestigial wings? 254 00:16:19,820 --> 00:16:21,940 Turns out they were normal. 255 00:16:21,940 --> 00:16:25,100 From that, what do we infer about these two traits, black 256 00:16:25,100 --> 00:16:27,780 body and vestigial wings? 257 00:16:27,780 --> 00:16:29,150 They're recessive traits. 258 00:16:29,150 --> 00:16:37,030 So here, the phenotype was normal appearance. 259 00:16:37,030 --> 00:16:42,950 So then he crosses them back, doing a test cross. 260 00:16:42,950 --> 00:16:46,240 Let's say he'll take males here and females here, but it 261 00:16:46,240 --> 00:16:48,030 actually works either way. 262 00:16:48,030 --> 00:16:50,440 And what he does is just like we did there. 263 00:16:50,440 --> 00:16:54,740 He could get gametes that were black, vestigial. 264 00:16:54,740 --> 00:17:00,480 He could get gametes that were black, plus; plus, vestigial; 265 00:17:00,480 --> 00:17:05,970 or plus, plus. 266 00:17:05,970 --> 00:17:09,210 Those are the four possibilities that come out. 267 00:17:09,210 --> 00:17:11,160 So when he does it-- let's keep score. 268 00:17:11,160 --> 00:17:12,488 I'm going to write them now-- 269 00:17:12,488 --> 00:17:18,390 plus, plus; black, vestigial. 270 00:17:18,390 --> 00:17:20,245 And from the other parent, you got this-- 271 00:17:23,540 --> 00:17:29,060 black, plus; black, vestigial; plus, 272 00:17:29,060 --> 00:17:32,310 vestigial; black, vestigial. 273 00:17:32,310 --> 00:17:34,280 Those are the four possibilities. 274 00:17:34,280 --> 00:17:37,460 And if this was just like Mendel's traits, it would be 275 00:17:37,460 --> 00:17:39,970 one to one to one to one. 276 00:17:39,970 --> 00:17:42,410 These were the parental types that went in. 277 00:17:42,410 --> 00:17:44,400 Plus, plus went in. 278 00:17:44,400 --> 00:17:46,130 And black, vestigial went in. 279 00:17:46,130 --> 00:17:48,690 Those were the combinations of traits here. 280 00:17:48,690 --> 00:17:52,490 These were new combinations. 281 00:17:52,490 --> 00:17:54,240 What he observed-- 282 00:17:54,240 --> 00:17:55,160 Let's see. 283 00:17:55,160 --> 00:17:57,610 If Mendel's right, it'll be one to one to one to one . 284 00:17:57,610 --> 00:17:59,510 If the chromosome theory's right, it'll be one to one to 285 00:17:59,510 --> 00:18:00,750 zero to zero. 286 00:18:00,750 --> 00:18:03,552 And who was right? 287 00:18:03,552 --> 00:18:04,450 STUDENT: No one. 288 00:18:04,450 --> 00:18:05,420 ERIC LANDER: No one. 289 00:18:05,420 --> 00:18:18,080 The answer was 965 to 944 to 206 to 185. 290 00:18:18,080 --> 00:18:21,540 Neither model was right. 291 00:18:21,540 --> 00:18:24,060 Neither model's right. 292 00:18:24,060 --> 00:18:27,180 The new combinations, the recombinant combinations, the 293 00:18:27,180 --> 00:18:29,490 non-parental combinations-- 294 00:18:29,490 --> 00:18:33,390 we can call these recombinant combinations. 295 00:18:33,390 --> 00:18:34,540 These were recombinant. 296 00:18:34,540 --> 00:18:36,350 They recombined in some way. 297 00:18:36,350 --> 00:18:38,390 They were a new combination, or they were the 298 00:18:38,390 --> 00:18:39,820 non-parental types. 299 00:18:39,820 --> 00:18:43,290 We use both of those words frequently-- 300 00:18:43,290 --> 00:18:47,990 were neither equal nor were they completely absent. 301 00:18:47,990 --> 00:18:50,850 They occurred, but at a lower frequency. 302 00:18:50,850 --> 00:18:51,950 What was the frequency? 303 00:18:51,950 --> 00:18:54,060 Well, we could just add it up. 304 00:18:54,060 --> 00:19:04,630 The frequency of recombinant types, of new types of were 305 00:19:04,630 --> 00:19:22,920 different than the parents', is 17%. 306 00:19:22,920 --> 00:19:26,280 What's going on? 307 00:19:26,280 --> 00:19:31,770 Now, maybe this is some magic number like the 3-to-1 ratio. 308 00:19:31,770 --> 00:19:34,820 And you should look at that 17% and say, ah, this is some 309 00:19:34,820 --> 00:19:37,840 constant of the universe, that when you put in traits you get 310 00:19:37,840 --> 00:19:41,650 17% percent of recombinant types. 311 00:19:41,650 --> 00:19:44,490 But it takes a little judgment to say, I don't think so. 312 00:19:44,490 --> 00:19:46,940 And he actually tried other traits. 313 00:19:46,940 --> 00:19:49,410 And sometimes he got one to one to one to one. 314 00:19:49,410 --> 00:19:53,780 But very often he got some funny number-- 315 00:19:53,780 --> 00:20:00,030 6%, 28%, 1%. 316 00:20:00,030 --> 00:20:03,090 There was some funny business going on. 317 00:20:03,090 --> 00:20:04,340 What's going on? 318 00:20:07,340 --> 00:20:09,110 Recombination. 319 00:20:09,110 --> 00:20:10,470 That's what's going on is there is 320 00:20:10,470 --> 00:20:12,290 recombination occurring. 321 00:20:12,290 --> 00:20:13,573 What do we mean by recombination? 322 00:20:16,890 --> 00:20:20,510 Recombination is very important stuff, by the way. 323 00:20:20,510 --> 00:20:23,740 At some point, I will tell you that understanding 324 00:20:23,740 --> 00:20:26,870 recombination was actually the origin of the 325 00:20:26,870 --> 00:20:28,770 Human Genome Project. 326 00:20:28,770 --> 00:20:31,210 And it traces back to a good MIT story. 327 00:20:31,210 --> 00:20:34,430 But that will be for a little later in the semester. 328 00:20:34,430 --> 00:20:37,010 So what do we think's happening here? 329 00:20:42,710 --> 00:20:46,560 I'm now going to draw a close-up of that chromosome. 330 00:20:46,560 --> 00:20:50,640 And here's another chromosome, the other pair. 331 00:20:56,320 --> 00:21:04,280 And what we think here might be happening is that you might 332 00:21:04,280 --> 00:21:16,380 have plus and black; black and plus-- 333 00:21:16,380 --> 00:21:17,700 oh, sorry. 334 00:21:17,700 --> 00:21:20,920 Black, right? 335 00:21:20,920 --> 00:21:22,316 Black, black; plus, plus. 336 00:21:22,316 --> 00:21:24,990 This would be plus. 337 00:21:24,990 --> 00:21:25,760 This would be plus. 338 00:21:25,760 --> 00:21:26,970 This is the normal chromosome. 339 00:21:26,970 --> 00:21:28,230 This is plus. 340 00:21:28,230 --> 00:21:34,000 So this chromosome here carries the pluses. 341 00:21:34,000 --> 00:21:39,880 This guy here carries those alleles black and vestigial. 342 00:21:43,640 --> 00:21:48,780 Well, what happens, the idea was that somehow these 343 00:21:48,780 --> 00:21:52,810 chromosomes exchanged material here. 344 00:21:52,810 --> 00:21:56,860 And the chromosome that was plus, plus; plus, plus now 345 00:21:56,860 --> 00:22:03,290 somehow acquires that bit, and this chromosome 346 00:22:03,290 --> 00:22:05,810 somehow gets that bit. 347 00:22:05,810 --> 00:22:19,510 And we end up instead with a picture like this, where some 348 00:22:19,510 --> 00:22:22,470 of this came from here. 349 00:22:27,730 --> 00:22:34,960 And those two loci, black and vestigial, were separated from 350 00:22:34,960 --> 00:22:38,350 each other such that the black allele moves over to that 351 00:22:38,350 --> 00:22:39,400 chromosome. 352 00:22:39,400 --> 00:22:42,200 Is that clear? 353 00:22:42,200 --> 00:22:43,300 That's the notion. 354 00:22:43,300 --> 00:22:45,510 Why did they think this was true? 355 00:22:45,510 --> 00:22:48,380 Well, it turns out that in fruit flies, you can actually 356 00:22:48,380 --> 00:22:50,250 look at eggs under the microscope. 357 00:22:50,250 --> 00:22:51,860 And you can look at the chromosomes. 358 00:22:51,860 --> 00:22:54,230 And if you take if you take the cells, and you take a 359 00:22:54,230 --> 00:22:57,010 cover slip, and you squash it down with your finger, you can 360 00:22:57,010 --> 00:23:02,180 actually see chromosomes lying right over each other, making 361 00:23:02,180 --> 00:23:05,250 little crosses like I drew there, up there, called 362 00:23:05,250 --> 00:23:07,650 chiasmata, which means crosses. 363 00:23:07,650 --> 00:23:11,310 And so people said, see, in the microscope, you can see 364 00:23:11,310 --> 00:23:13,990 they're lying on top of each other. 365 00:23:13,990 --> 00:23:15,540 Are you impressed by that piece of evidence? 366 00:23:18,800 --> 00:23:19,140 No. 367 00:23:19,140 --> 00:23:20,640 You took the cover slip, you squished it 368 00:23:20,640 --> 00:23:21,470 down with your fingers. 369 00:23:21,470 --> 00:23:23,400 So they're lying on top of each other? 370 00:23:23,400 --> 00:23:25,450 Big deal. 371 00:23:25,450 --> 00:23:26,480 I'm not going to be impressed. 372 00:23:26,480 --> 00:23:28,890 And calling it chiasmata doesn't make me any more 373 00:23:28,890 --> 00:23:30,280 impressed, right? 374 00:23:30,280 --> 00:23:32,430 Although it's always good to call things Greek names, 375 00:23:32,430 --> 00:23:34,850 because people think they're more sophisticated if you call 376 00:23:34,850 --> 00:23:37,920 them Greek names. 377 00:23:37,920 --> 00:23:40,490 But this was the notion people had. 378 00:23:40,490 --> 00:23:46,600 And the frequency 17% would indicate how often these 379 00:23:46,600 --> 00:23:47,850 crossovers occur. 380 00:23:51,280 --> 00:23:53,750 But if I were in the situation we talked about with Mendel, 381 00:23:53,750 --> 00:23:56,460 and I wrote this up, and I said, see, it's 17% sometimes, 382 00:23:56,460 --> 00:23:58,610 6% sometimes, 28% sometimes; and when I look in the 383 00:23:58,610 --> 00:24:01,440 microscope, they lie on top of each other; therefore, it's 384 00:24:01,440 --> 00:24:02,690 recombination-- 385 00:24:05,180 --> 00:24:07,310 There were actually other ideas floating around too. 386 00:24:07,310 --> 00:24:11,580 Maybe it has something to do with developmental biology. 387 00:24:11,580 --> 00:24:13,820 It was a puzzle. 388 00:24:13,820 --> 00:24:18,370 When you have a really deep puzzle, the most important 389 00:24:18,370 --> 00:24:22,890 thing in science is to find a young person. 390 00:24:22,890 --> 00:24:25,960 Because young people come without prejudice. 391 00:24:25,960 --> 00:24:27,580 They say, let me just look at all of this. 392 00:24:27,580 --> 00:24:28,200 Stand back. 393 00:24:28,200 --> 00:24:29,610 I don't come with any prejudice. 394 00:24:29,610 --> 00:24:32,310 So at MIT, what is the solution when you have a 395 00:24:32,310 --> 00:24:34,270 problem like this? 396 00:24:34,270 --> 00:24:35,460 A UROP. 397 00:24:35,460 --> 00:24:37,840 You want a UROP. 398 00:24:37,840 --> 00:24:41,405 So even 1911, that was the solution at Columbia. 399 00:24:44,300 --> 00:24:47,970 Thomas Hunt Morgan got a UROP. 400 00:24:47,970 --> 00:24:48,480 I'm serious. 401 00:24:48,480 --> 00:24:51,200 He was called Alfred Sturtevant. 402 00:24:51,200 --> 00:24:55,950 Alfred Sturtevant was a sophomore at Columbia. 403 00:24:55,950 --> 00:24:58,570 Everybody else was busy finding this recombination 404 00:24:58,570 --> 00:25:01,482 data, how often this recombined with this, this 405 00:25:01,482 --> 00:25:02,780 with this, this with this. 406 00:25:02,780 --> 00:25:04,400 Sturtevant-- 407 00:25:04,400 --> 00:25:05,680 he's a sophomore-- 408 00:25:05,680 --> 00:25:09,050 he says, god, this stuff's interesting! 409 00:25:09,050 --> 00:25:11,690 Professor Morgan, could I have all the data and 410 00:25:11,690 --> 00:25:13,800 try to look at it? 411 00:25:13,800 --> 00:25:17,760 Sturtevant took it home and actually pulled an 412 00:25:17,760 --> 00:25:20,810 all-nighter, blew off his homework-- 413 00:25:20,810 --> 00:25:22,920 it actually says so in his autobiography. 414 00:25:22,920 --> 00:25:25,660 He says, I blew off my homework and pulled an 415 00:25:25,660 --> 00:25:27,370 all-nighter-- 416 00:25:27,370 --> 00:25:30,140 essentially in those words, he says. "To the detriment of my 417 00:25:30,140 --> 00:25:33,140 undergraduate homework" is the way he puts it. 418 00:25:33,140 --> 00:25:42,310 But in any case, genetic maps and Sturtevant's all-nighter. 419 00:25:45,730 --> 00:25:47,970 What Sturtevant did was the following. 420 00:25:47,970 --> 00:25:54,300 He said, how are we going to prove the chromosome theory? 421 00:25:54,300 --> 00:25:57,850 I like this idea that recombination is about 422 00:25:57,850 --> 00:25:59,950 distance on the chromosomes. 423 00:25:59,950 --> 00:26:04,960 I like the concept that maybe 17% is how often 424 00:26:04,960 --> 00:26:07,800 these things recombine. 425 00:26:07,800 --> 00:26:10,770 Why would things only recombine 1% of the time? 426 00:26:10,770 --> 00:26:12,020 What would that mean? 427 00:26:14,092 --> 00:26:16,400 They've got to be pretty close together so that a crossover 428 00:26:16,400 --> 00:26:19,130 between them happens not so often. 429 00:26:19,130 --> 00:26:22,660 And what if things were far apart? 430 00:26:22,660 --> 00:26:23,920 Well, it could be more likely. 431 00:26:23,920 --> 00:26:26,310 So he likes the idea that recombination 432 00:26:26,310 --> 00:26:29,175 frequency means distance. 433 00:26:29,175 --> 00:26:31,230 But how are you going to prove that? 434 00:26:31,230 --> 00:26:33,290 It could mean a zillion other things. 435 00:26:33,290 --> 00:26:35,090 It could mean biochemical pathways, 436 00:26:35,090 --> 00:26:36,930 developmental biology. 437 00:26:36,930 --> 00:26:39,610 How are you going to prove that recombination frequency 438 00:26:39,610 --> 00:26:41,810 means distance? 439 00:26:41,810 --> 00:26:43,500 You've got to make predictions, right? 440 00:26:43,500 --> 00:26:46,680 The only to do it would be to make predictions. 441 00:26:46,680 --> 00:26:51,960 So Sturtevant takes the data, and he starts making 442 00:26:51,960 --> 00:26:53,210 predictions. 443 00:26:58,982 --> 00:27:04,150 I think, says Sturtevant, these things are alleles 444 00:27:04,150 --> 00:27:09,020 living at genes with locations on the chromosome-- 445 00:27:09,020 --> 00:27:11,480 black, vestigial. 446 00:27:14,590 --> 00:27:18,300 How often do black and vestigial 447 00:27:18,300 --> 00:27:19,350 recombine with each other? 448 00:27:19,350 --> 00:27:20,940 What is the frequency of recombinant 449 00:27:20,940 --> 00:27:23,050 non-parental types? 450 00:27:23,050 --> 00:27:26,360 17%. 451 00:27:26,360 --> 00:27:30,860 So Sturtevant goes through the data and he says, what about 452 00:27:30,860 --> 00:27:33,265 other crosses people did in the lab? 453 00:27:33,265 --> 00:27:36,770 Well, it turns out people did crosses with another mutant 454 00:27:36,770 --> 00:27:41,730 that produces a funny eye color called cinnabar, Cn. 455 00:27:41,730 --> 00:27:44,020 So cinnabar. 456 00:27:47,470 --> 00:27:55,010 It turns out that the recombination frequency 457 00:27:55,010 --> 00:28:07,310 between cinnabar to vestigial was 8%. 458 00:28:07,310 --> 00:28:12,620 Vestigial, cinnabar, 8% recombination. 459 00:28:12,620 --> 00:28:14,640 If this chromosome business is right, where 460 00:28:14,640 --> 00:28:15,890 should I put cinnabar? 461 00:28:18,140 --> 00:28:19,160 Sorry? 462 00:28:19,160 --> 00:28:21,084 Where do you want it? 463 00:28:21,084 --> 00:28:21,900 STUDENT: [INAUDIBLE]. 464 00:28:21,900 --> 00:28:22,920 ERIC LANDER: Somewhere in between. 465 00:28:22,920 --> 00:28:26,570 You'd like me to put that there. 466 00:28:26,570 --> 00:28:27,920 STUDENT: On the other side. 467 00:28:27,920 --> 00:28:28,420 ERIC LANDER: Oh, OK. 468 00:28:28,420 --> 00:28:28,940 Wait a second. 469 00:28:28,940 --> 00:28:31,361 On the other side. 470 00:28:31,361 --> 00:28:32,430 OK, which is it? 471 00:28:32,430 --> 00:28:34,765 How many vote for the left? 472 00:28:34,765 --> 00:28:37,080 How many vote for the right? 473 00:28:37,080 --> 00:28:40,510 How many conscientious abstainers are there? 474 00:28:40,510 --> 00:28:41,930 Do we know? 475 00:28:41,930 --> 00:28:42,850 STUDENT: No. 476 00:28:42,850 --> 00:28:43,180 ERIC LANDER: No. 477 00:28:43,180 --> 00:28:44,880 There are two possibilities. 478 00:28:44,880 --> 00:28:52,160 It could be 8% this way, or it could be 8% that way. 479 00:28:52,160 --> 00:28:53,410 How are we going to know? 480 00:28:53,410 --> 00:28:55,223 Yes? 481 00:28:55,223 --> 00:28:57,150 STUDENT: [INAUDIBLE]. 482 00:28:57,150 --> 00:28:57,990 ERIC LANDER: Black. 483 00:28:57,990 --> 00:29:02,750 What if we knew the answer between cinnabar and black? 484 00:29:02,750 --> 00:29:04,340 That would constrain the problem. 485 00:29:04,340 --> 00:29:06,040 Can you give me two predictions for what the 486 00:29:06,040 --> 00:29:07,282 answer might be? 487 00:29:07,282 --> 00:29:08,520 STUDENT: Either 9% or-- 488 00:29:08,520 --> 00:29:12,310 ERIC LANDER: Either 9% or 25%. 489 00:29:12,310 --> 00:29:13,430 So now we have a prediction. 490 00:29:13,430 --> 00:29:15,570 We don't know where cinnabar is. 491 00:29:15,570 --> 00:29:20,580 But the answer could be that black to vestigial should 492 00:29:20,580 --> 00:29:23,990 either be about 9%-- 493 00:29:23,990 --> 00:29:25,660 that's what that would be here-- 494 00:29:25,660 --> 00:29:32,140 or about 25%. 495 00:29:32,140 --> 00:29:33,390 The answer? 496 00:29:36,020 --> 00:29:37,270 About 9%. 497 00:29:40,860 --> 00:29:42,430 That's what Sturtevant found. 498 00:29:46,550 --> 00:29:49,900 That's a prediction and kind of cool. 499 00:29:49,900 --> 00:29:53,230 And you can imagine taking the data home and it's probably 9 500 00:29:53,230 --> 00:29:55,780 o'clock, and you've now realized, wow, 501 00:29:55,780 --> 00:29:58,320 freaky, it's 9%. 502 00:29:58,320 --> 00:30:04,430 So then he looked at the mutation "lobe." Lobe was 503 00:30:04,430 --> 00:30:06,440 another mutation. 504 00:30:06,440 --> 00:30:12,530 The lobe mutation showed 5% recombination from vestigial. 505 00:30:15,950 --> 00:30:17,200 Where should we put it? 506 00:30:20,190 --> 00:30:24,230 Left or right, well, let's make some predictions. 507 00:30:27,810 --> 00:30:31,220 Suppose it's over here. 508 00:30:31,220 --> 00:30:34,360 Will it be very close to cinnabar? 509 00:30:34,360 --> 00:30:37,025 Will it be closer to black? 510 00:30:37,025 --> 00:30:39,990 But what if it was over here? 511 00:30:39,990 --> 00:30:41,890 Well, it would be further. 512 00:30:41,890 --> 00:30:45,940 So let's put lobe in. 513 00:30:45,940 --> 00:30:50,960 And suppose we know that lob is 5%. 514 00:30:50,960 --> 00:30:54,665 Then what's the prediction for black to lobe? 515 00:30:58,890 --> 00:31:01,660 22%. 516 00:31:01,660 --> 00:31:09,090 Answer, according to the notebooks, 21%, pretty close. 517 00:31:09,090 --> 00:31:10,550 You'd like it to be exactly 22. 518 00:31:10,550 --> 00:31:12,020 But life doesn't always turn out that way. 519 00:31:12,020 --> 00:31:15,760 21's pretty close to 22. 520 00:31:15,760 --> 00:31:17,150 What other predictions could we make? 521 00:31:17,150 --> 00:31:19,540 If this is cinnabar here, could you give me a prediction 522 00:31:19,540 --> 00:31:23,620 for cinnabar to lobe? 523 00:31:23,620 --> 00:31:25,450 13%. 524 00:31:25,450 --> 00:31:27,540 Yep, that works. 525 00:31:27,540 --> 00:31:28,790 Curved wing. 526 00:31:33,470 --> 00:31:38,640 Recombination distance from lobed, 3%. 527 00:31:38,640 --> 00:31:39,950 So now you have some predictions. 528 00:31:39,950 --> 00:31:41,460 You have this prediction here. 529 00:31:41,460 --> 00:31:43,180 You predict 8%. 530 00:31:43,180 --> 00:31:45,250 Answer, about 8%. 531 00:31:45,250 --> 00:31:50,660 Over here you predict 16%. 532 00:31:50,660 --> 00:31:53,730 Answer, about 16%. 533 00:31:53,730 --> 00:31:56,100 Bingo. 534 00:31:56,100 --> 00:32:00,960 Sturtivant says, if these genes-- 535 00:32:00,960 --> 00:32:02,670 we call them loci, often. 536 00:32:02,670 --> 00:32:05,610 I'll use the word locus synonymous with gene. 537 00:32:05,610 --> 00:32:07,370 Locus means a place. 538 00:32:07,370 --> 00:32:09,820 And geneticists think about genes as a place on a 539 00:32:09,820 --> 00:32:10,540 chromosome. 540 00:32:10,540 --> 00:32:12,320 If these loci-- 541 00:32:12,320 --> 00:32:16,570 the plural of locus-- if these loci were really arrayed along 542 00:32:16,570 --> 00:32:21,760 a linear structure, then it would have to be the case that 543 00:32:21,760 --> 00:32:23,380 they would have certain additive 544 00:32:23,380 --> 00:32:25,750 relationships between them. 545 00:32:25,750 --> 00:32:27,810 And the chance that they would have these additive 546 00:32:27,810 --> 00:32:31,900 relationships if they weren't part of a line is pretty 547 00:32:31,900 --> 00:32:33,820 implausible. 548 00:32:33,820 --> 00:32:39,130 That's a real prediction, a very remarkable prediction. 549 00:32:39,130 --> 00:32:41,420 And it holds up with the data. 550 00:32:41,420 --> 00:32:43,260 Sturtivant pulls the all-nighter. 551 00:32:43,260 --> 00:32:46,090 By the time the sun comes up at Columbia University-- this 552 00:32:46,090 --> 00:32:47,700 is Morningside Heights-- 553 00:32:47,700 --> 00:32:50,360 he's got the whole thing worked out. 554 00:32:50,360 --> 00:32:53,340 Yes, this chromosome theory must be right. 555 00:32:53,340 --> 00:32:57,520 It fits beautifully all of these data. 556 00:32:57,520 --> 00:32:59,760 Pretty cool. 557 00:32:59,760 --> 00:33:04,220 You are all authorized to blow off your homework anytime you 558 00:33:04,220 --> 00:33:07,032 make a discovery like that. 559 00:33:07,032 --> 00:33:08,740 [LAUGHTER] 560 00:33:08,740 --> 00:33:10,030 That's a course rule. 561 00:33:10,030 --> 00:33:13,230 Any homework will be forgiven for 562 00:33:13,230 --> 00:33:17,080 discoveries of that magnitude. 563 00:33:17,080 --> 00:33:19,810 All right. 564 00:33:19,810 --> 00:33:21,060 Tell your TAs. 565 00:33:23,500 --> 00:33:25,770 So now, what does it really tell us? 566 00:33:25,770 --> 00:33:30,565 It tells us that if genes are very close together, the 567 00:33:30,565 --> 00:33:35,730 recombination frequency, R recombination frequency, or 568 00:33:35,730 --> 00:33:38,260 RF, might be very little. 569 00:33:38,260 --> 00:33:40,660 They could be as low as almost zero, which means you never 570 00:33:40,660 --> 00:33:42,760 see a recombinant because they're right 571 00:33:42,760 --> 00:33:44,130 next to each other. 572 00:33:44,130 --> 00:33:47,800 Or it could be that they're further apart. 573 00:33:47,800 --> 00:33:48,990 It might be 1%. 574 00:33:48,990 --> 00:33:50,950 It might be 10%. 575 00:33:50,950 --> 00:33:52,100 It could keep growing. 576 00:33:52,100 --> 00:33:55,250 It might be 30%. 577 00:33:55,250 --> 00:34:00,100 Suppose it's way, way, way, way, way far away. 578 00:34:00,100 --> 00:34:02,360 What's the largest it ever gets to be? 579 00:34:06,150 --> 00:34:08,020 Well, if they were on different chromosomes-- 580 00:34:08,020 --> 00:34:10,310 suppose there were totally independent segregation, 581 00:34:10,310 --> 00:34:11,889 different chromosomes-- 582 00:34:11,889 --> 00:34:14,790 what would they be? 583 00:34:14,790 --> 00:34:15,238 STUDENT: [INAUDIBLE] 584 00:34:15,238 --> 00:34:17,982 ERIC LANDER: No, it's 50% because remember, one to one 585 00:34:17,982 --> 00:34:19,969 to one to one says that the recombinant-- 586 00:34:19,969 --> 00:34:21,090 well, actually, this is an interesting question. 587 00:34:21,090 --> 00:34:22,000 Let's come to that 100. 588 00:34:22,000 --> 00:34:25,300 On different chromosomes, one to one to one to one means 589 00:34:25,300 --> 00:34:28,489 that it's 50%, half of them are recombinant types. 590 00:34:28,489 --> 00:34:31,469 It turns out on the same chromosome, as you get farther 591 00:34:31,469 --> 00:34:33,730 and farther and farther, you might say there's going to be 592 00:34:33,730 --> 00:34:37,290 100% chance of a crossover. 593 00:34:37,290 --> 00:34:39,110 And you might say the recombination frequency could 594 00:34:39,110 --> 00:34:42,010 keep growing past 50%. 595 00:34:42,010 --> 00:34:43,380 It turns out it doesn't. 596 00:34:43,380 --> 00:34:49,150 The reason is that multiple crossovers can happen. 597 00:34:49,150 --> 00:34:53,060 So mathematical interjection here, if here's my gene and 598 00:34:53,060 --> 00:34:58,760 here's my gene, there could be one crossover. 599 00:34:58,760 --> 00:35:01,400 There could be two crossovers. 600 00:35:01,400 --> 00:35:05,110 There could be three crossovers. 601 00:35:05,110 --> 00:35:09,620 And so in fact, it turns out, as you get very far away, you 602 00:35:09,620 --> 00:35:12,390 have to start paying attention to the probabilities of double 603 00:35:12,390 --> 00:35:15,680 crossovers and triple crossovers. 604 00:35:15,680 --> 00:35:18,560 And so it turns out that it's a Poisson process of the 605 00:35:18,560 --> 00:35:21,240 number of crossovers that occurs, give or take. 606 00:35:21,240 --> 00:35:25,490 And you see a recombination if there's an odd number. 607 00:35:25,490 --> 00:35:28,470 And for that reason, it never gets above 50%. 608 00:35:28,470 --> 00:35:30,680 So as the distance gets further and further and 609 00:35:30,680 --> 00:35:33,740 further, it goes from zero to 50, which is the same number 610 00:35:33,740 --> 00:35:35,760 you get for separate chromosomes. 611 00:35:35,760 --> 00:35:37,620 Otherwise, you might think that if there was just one 612 00:35:37,620 --> 00:35:40,540 crossover, it gets to 100% probability of recombination. 613 00:35:40,540 --> 00:35:42,360 But it never does, because there are doubles. 614 00:35:42,360 --> 00:35:44,960 And you can actually observe, if you make a cross that has 615 00:35:44,960 --> 00:35:47,560 three different genes segregating in it, you can 616 00:35:47,560 --> 00:35:51,380 actually see the double crossover type. 617 00:35:51,380 --> 00:35:55,400 So you can see very nicely that if you make a cross 618 00:35:55,400 --> 00:36:03,650 involving black, cinnabar, and vestigial over plus, plus, 619 00:36:03,650 --> 00:36:06,710 plus, you can see that cinnabar is right in the 620 00:36:06,710 --> 00:36:10,870 middle because this recombination happens at a 621 00:36:10,870 --> 00:36:12,930 pretty good frequency. 622 00:36:12,930 --> 00:36:17,430 This recombination happens at a good frequency, giving you 623 00:36:17,430 --> 00:36:21,860 plus, cinnabar, vestigial or black, plus, plus. 624 00:36:21,860 --> 00:36:26,060 Sometimes you will get that recombination. 625 00:36:26,060 --> 00:36:30,320 You'll get out gametes that are black, plus, vestigial, 626 00:36:30,320 --> 00:36:33,280 but at a much lower frequency because they take two 627 00:36:33,280 --> 00:36:34,230 crossovers. 628 00:36:34,230 --> 00:36:36,490 What will be the probability of seeing a 629 00:36:36,490 --> 00:36:38,600 black, cinnabar, vestigial? 630 00:36:38,600 --> 00:36:42,940 Well, we said that this one was about 9%, this 631 00:36:42,940 --> 00:36:45,040 one was about 8%. 632 00:36:45,040 --> 00:36:48,980 What's the product of a 9% chance and an 8% chance? 633 00:36:48,980 --> 00:36:52,000 It's about a 1% chance, a little less than a 1% chance. 634 00:36:52,000 --> 00:36:55,520 That how frequently you see black, plus, vestigial. 635 00:36:55,520 --> 00:36:58,170 You can even predict the probability of a double 636 00:36:58,170 --> 00:37:01,160 crossover event by multiplying the two events 637 00:37:01,160 --> 00:37:02,660 that have to happen. 638 00:37:02,660 --> 00:37:05,940 So your bottom-line rules here is that because of these 639 00:37:05,940 --> 00:37:09,410 double crossovers our recombination frequency can go 640 00:37:09,410 --> 00:37:15,640 from zero to about 50%. 641 00:37:15,640 --> 00:37:19,910 This is independent assortment. 642 00:37:19,910 --> 00:37:25,650 And it either occurs if you're on different chromosomes or, 643 00:37:25,650 --> 00:37:29,015 if you're very far away on the same chromosome, they behave 644 00:37:29,015 --> 00:37:31,260 as if they're independent of each other. 645 00:37:31,260 --> 00:37:33,570 Any questions about any of that? 646 00:37:33,570 --> 00:37:34,563 Yes? 647 00:37:34,563 --> 00:37:36,535 STUDENT: Why didn't Mendel ever see recombination? 648 00:37:36,535 --> 00:37:38,470 ERIC LANDER: Why didn't Mendel ever see recombination? 649 00:37:38,470 --> 00:37:40,710 It turns out that with seven chromosomes, and they're 650 00:37:40,710 --> 00:37:44,380 biggish in length, he never actually ran into two loci 651 00:37:44,380 --> 00:37:47,150 that were close enough, to notice it. 652 00:37:47,150 --> 00:37:48,170 That's why. 653 00:37:48,170 --> 00:37:52,320 Flies actually only have three major chromosomes. 654 00:37:52,320 --> 00:37:55,440 There's a fourth, but it's a puny little thing. 655 00:37:55,440 --> 00:37:58,440 And because they were much more intensively collecting 656 00:37:58,440 --> 00:38:01,810 mutations in Morgan's fly room, they began having a lot 657 00:38:01,810 --> 00:38:03,110 of them, and they had to bump into 658 00:38:03,110 --> 00:38:05,960 recombination pretty early. 659 00:38:05,960 --> 00:38:08,670 Mendel simply didn't have enough that were close enough. 660 00:38:08,670 --> 00:38:10,770 Think about what would have happened if just by chance, 661 00:38:10,770 --> 00:38:14,790 somebody were selling a strain of peas in the market which 662 00:38:14,790 --> 00:38:19,310 had a mutation in a locus that had 10% recombination 663 00:38:19,310 --> 00:38:22,200 distance, and it screwed up Mendel's law of independent 664 00:38:22,200 --> 00:38:24,560 assortment of two loci? 665 00:38:24,560 --> 00:38:26,720 Mendel might not have published the paper. 666 00:38:26,720 --> 00:38:30,450 Sometimes in science it's actually valuable to first get 667 00:38:30,450 --> 00:38:34,520 the oversimplification out , like his second law, and then 668 00:38:34,520 --> 00:38:36,650 deal with the complexity that sits on top of the 669 00:38:36,650 --> 00:38:38,030 oversimplification. 670 00:38:38,030 --> 00:38:40,810 It's kind of lucky that Mendel didn't have enough of them to 671 00:38:40,810 --> 00:38:43,620 be bothered, in the first paper. 672 00:38:43,620 --> 00:38:45,680 So in fact, all of Mendel's seven loci 673 00:38:45,680 --> 00:38:46,440 have now been mapped. 674 00:38:46,440 --> 00:38:48,120 Most have been cloned molecularly. 675 00:38:48,120 --> 00:38:50,470 And so we actually know where they are, et cetera. 676 00:38:50,470 --> 00:38:51,790 It's a really good question. 677 00:38:51,790 --> 00:38:53,240 That's sort of why he didn't.