0 00:00:00,000 --> 00:00:30,000 Dear viewer, these subtitles were generated by a machine via the service Trint and therefore are (very) buggy. If you are capable, please help us to create good quality subtitles: https://c3subtitles.de/talk/910 Thanks! 1 00:00:14,690 --> 00:00:16,369 And with that being said, I'd like to 2 00:00:16,370 --> 00:00:18,319 introduce our next speaker for this 3 00:00:18,320 --> 00:00:19,729 evening. 4 00:00:19,730 --> 00:00:22,339 So she is a medical biotechnologist 5 00:00:22,340 --> 00:00:23,239 with a Ph.D. 6 00:00:23,240 --> 00:00:25,369 in cancer research and 7 00:00:25,370 --> 00:00:28,399 she'll be talking today about extending 8 00:00:28,400 --> 00:00:29,779 DNA analysis. 9 00:00:29,780 --> 00:00:30,809 So please welcome. 10 00:00:30,810 --> 00:00:31,810 And I. Molner. 11 00:00:38,950 --> 00:00:41,989 Well, thank you for being here and. 12 00:00:41,990 --> 00:00:44,479 Yeah. And for the Cannes production. 13 00:00:44,480 --> 00:00:47,709 So just a quick recap. 14 00:00:47,710 --> 00:00:50,409 I'm a biologist by training and 15 00:00:50,410 --> 00:00:52,129 did my part in cancer research. 16 00:00:52,130 --> 00:00:54,549 Now I work as a PR consultant 17 00:00:54,550 --> 00:00:56,949 for health care as my day job. 18 00:00:56,950 --> 00:00:59,199 And I'm 19 00:00:59,200 --> 00:01:01,329 also a science communicator. 20 00:01:01,330 --> 00:01:03,669 And this is seen as something 21 00:01:03,670 --> 00:01:05,169 of science. Communication has nothing to 22 00:01:05,170 --> 00:01:06,640 do with my day job. 23 00:01:08,380 --> 00:01:11,019 And I wish to talk to you about DNA. 24 00:01:11,020 --> 00:01:13,749 You've probably heard of it and 25 00:01:13,750 --> 00:01:16,059 it stores our genetic information. 26 00:01:16,060 --> 00:01:17,919 So it's a different kind of code than 27 00:01:17,920 --> 00:01:19,329 most people here are used to. 28 00:01:20,590 --> 00:01:22,689 It's made up of four bases 29 00:01:22,690 --> 00:01:25,659 adenine, thymine, guanine and cytosine, 30 00:01:25,660 --> 00:01:27,839 which are usually abbreviated 80 31 00:01:27,840 --> 00:01:29,109 GNC. 32 00:01:29,110 --> 00:01:31,479 And as you can see in the image, 33 00:01:31,480 --> 00:01:33,609 the adenine always finds 34 00:01:33,610 --> 00:01:35,809 thymine and guanine always bind 35 00:01:35,810 --> 00:01:38,529 cytosine and vice versa. 36 00:01:38,530 --> 00:01:40,899 And so you have two 37 00:01:40,900 --> 00:01:42,999 negative of one positive code 38 00:01:43,000 --> 00:01:44,079 always. 39 00:01:44,080 --> 00:01:46,299 And there's coding and non 40 00:01:46,300 --> 00:01:47,979 coding DNA. 41 00:01:47,980 --> 00:01:50,109 So the coding DNA is actually 42 00:01:50,110 --> 00:01:52,239 genes. So everything that shows up 43 00:01:52,240 --> 00:01:54,399 in your looks and the non 44 00:01:54,400 --> 00:01:57,039 coding DNA is not genes. 45 00:01:57,040 --> 00:01:59,559 So everything outside of your genes, 46 00:01:59,560 --> 00:02:01,929 both are important. 47 00:02:01,930 --> 00:02:03,809 And DNA. 48 00:02:03,810 --> 00:02:06,189 You might have heard is also used as 49 00:02:06,190 --> 00:02:08,379 evidence in crimes, 50 00:02:08,380 --> 00:02:10,779 whereas ninety nine point nine percent 51 00:02:10,780 --> 00:02:12,999 of our DNA is the same, which is a good 52 00:02:13,000 --> 00:02:15,069 thing because otherwise we would all be 53 00:02:15,070 --> 00:02:16,070 mutants. 54 00:02:17,290 --> 00:02:19,869 0.01 percent is different 55 00:02:19,870 --> 00:02:21,999 and differs from person to person. 56 00:02:22,000 --> 00:02:24,669 And in 1984, Alec Jeffries 57 00:02:24,670 --> 00:02:27,069 discovered DNA profiling 58 00:02:27,070 --> 00:02:29,559 where he saw that he could use this 0.01 59 00:02:29,560 --> 00:02:31,089 percent to 60 00:02:32,230 --> 00:02:34,029 identify people. 61 00:02:34,030 --> 00:02:36,399 And this is possible to compare 62 00:02:36,400 --> 00:02:38,679 the DNA profile of one person 63 00:02:38,680 --> 00:02:40,959 with another person and can 64 00:02:40,960 --> 00:02:43,389 also be used to find relatives 65 00:02:43,390 --> 00:02:45,849 for paternity issues, for example. 66 00:02:45,850 --> 00:02:47,949 And often this is 67 00:02:47,950 --> 00:02:50,529 referred to as DNA fingerprints. 68 00:02:50,530 --> 00:02:52,659 And what this really means is checking 69 00:02:52,660 --> 00:02:55,119 for short tandem repeats 70 00:02:55,120 --> 00:02:56,889 or S.T. cars. 71 00:02:56,890 --> 00:02:58,959 And these Astaire's are repetitive 72 00:02:58,960 --> 00:03:01,179 sequences outside of 73 00:03:01,180 --> 00:03:02,229 genes. 74 00:03:02,230 --> 00:03:03,969 So they have nothing to do with how we 75 00:03:03,970 --> 00:03:04,970 look like. 76 00:03:06,120 --> 00:03:08,289 And they are many STF yars with 77 00:03:08,290 --> 00:03:10,929 different lengths which are inherited. 78 00:03:10,930 --> 00:03:13,299 And the pattern of these SDR 79 00:03:13,300 --> 00:03:15,279 differs from person to person. 80 00:03:15,280 --> 00:03:17,379 And basically, it just looks like this 81 00:03:17,380 --> 00:03:19,209 that you have, for example, the repeat 82 00:03:19,210 --> 00:03:21,339 CTA five times 83 00:03:21,340 --> 00:03:24,159 or six times or nine times. 84 00:03:24,160 --> 00:03:26,529 And the number 85 00:03:26,530 --> 00:03:28,779 of the repeats is the thing that 86 00:03:28,780 --> 00:03:29,780 differs. 87 00:03:31,630 --> 00:03:33,759 If you look at the gene, then 88 00:03:33,760 --> 00:03:35,889 this could look something like this. 89 00:03:35,890 --> 00:03:38,049 And basically you just count 90 00:03:38,050 --> 00:03:39,589 the number of the repeat. 91 00:03:39,590 --> 00:03:42,339 So each box here represents a repeat. 92 00:03:42,340 --> 00:03:44,589 Sometimes you have a similar one 93 00:03:44,590 --> 00:03:46,809 and the same number and sometimes 94 00:03:46,810 --> 00:03:47,860 you have different numbers. 95 00:03:49,060 --> 00:03:51,189 What do you do to get a DNA fingerprint 96 00:03:51,190 --> 00:03:53,589 as first that you isolate the DNA from 97 00:03:53,590 --> 00:03:55,839 a crime scene or form a probe? 98 00:03:55,840 --> 00:03:57,399 You amplify these S.T. 99 00:03:57,400 --> 00:03:59,529 eyes and then you can separate 100 00:03:59,530 --> 00:04:01,749 them by their size because the different 101 00:04:01,750 --> 00:04:04,149 number of repeats generates a different 102 00:04:04,150 --> 00:04:06,789 size of the amplified 103 00:04:06,790 --> 00:04:08,409 sequence. 104 00:04:08,410 --> 00:04:10,519 And this you do 216 105 00:04:10,520 --> 00:04:11,979 S.T. Rs. 106 00:04:11,980 --> 00:04:13,689 And then you get a profile that looks 107 00:04:13,690 --> 00:04:15,519 something like this, which you might have 108 00:04:15,520 --> 00:04:18,069 seen in a TV series already. 109 00:04:18,070 --> 00:04:20,229 And what 110 00:04:20,230 --> 00:04:23,019 I said was that the DNA fingerprints 111 00:04:23,020 --> 00:04:25,509 don't really show how you look like. 112 00:04:25,510 --> 00:04:27,100 However, in. 113 00:04:28,630 --> 00:04:30,579 Because they are outside of genes. 114 00:04:30,580 --> 00:04:32,829 However, as you may 115 00:04:32,830 --> 00:04:35,199 remember, we have 116 00:04:35,200 --> 00:04:37,299 chromosomes. And on these two 117 00:04:37,300 --> 00:04:38,979 of these chromosomes are the so-called 118 00:04:38,980 --> 00:04:40,689 sex chromosomes. 119 00:04:40,690 --> 00:04:43,699 And these are called X and Y. 120 00:04:43,700 --> 00:04:44,700 And 121 00:04:45,850 --> 00:04:48,669 you can identify the X chromosomes 122 00:04:48,670 --> 00:04:51,159 and the Y chromosomes by the gene 123 00:04:51,160 --> 00:04:53,439 of the enamel of our tooth 124 00:04:53,440 --> 00:04:54,429 enamel. 125 00:04:54,430 --> 00:04:57,009 And since 2004, 126 00:04:57,010 --> 00:04:59,079 we can use this to identify the 127 00:04:59,080 --> 00:05:01,239 sex chromosomes of a person 128 00:05:01,240 --> 00:05:02,739 because the X chromosome 129 00:05:03,760 --> 00:05:06,399 gene is six paces shorter. 130 00:05:06,400 --> 00:05:08,470 But this is not always accurate. 131 00:05:09,700 --> 00:05:12,319 So this is how the DNA fingerprint 132 00:05:12,320 --> 00:05:14,079 then looks like in the. 133 00:05:15,850 --> 00:05:17,079 For the scientists. 134 00:05:17,080 --> 00:05:18,459 This is an electrified program. 135 00:05:18,460 --> 00:05:20,709 Basically just a measurement off the bay. 136 00:05:20,710 --> 00:05:22,689 A band that you just saw. 137 00:05:22,690 --> 00:05:24,819 And you can also see on the 138 00:05:24,820 --> 00:05:26,979 up on the lower left, the 139 00:05:26,980 --> 00:05:28,660 X chromosome peak. 140 00:05:29,770 --> 00:05:31,479 And you see there is no Y chromosome 141 00:05:31,480 --> 00:05:33,519 peak. So this person has two X 142 00:05:33,520 --> 00:05:35,199 chromosomes. 143 00:05:35,200 --> 00:05:37,179 And this is just a means of identity. 144 00:05:37,180 --> 00:05:38,379 Vacation. 145 00:05:38,380 --> 00:05:40,629 So it does not show 146 00:05:40,630 --> 00:05:42,909 much about the person besides the sex 147 00:05:42,910 --> 00:05:45,219 chromosomes where you can determine 148 00:05:45,220 --> 00:05:46,689 the sex of the person. 149 00:05:48,490 --> 00:05:50,619 And in two 150 00:05:50,620 --> 00:05:52,719 thousand and sixteen, there was 151 00:05:52,720 --> 00:05:54,909 a rape case in Freiburg 152 00:05:54,910 --> 00:05:57,669 and actually rape, murder, where Maria 153 00:05:57,670 --> 00:05:59,649 was found raped and murdered at the drys 154 00:05:59,650 --> 00:06:01,799 arm of a river in Baton Rouge 155 00:06:01,800 --> 00:06:03,219 and back. 156 00:06:03,220 --> 00:06:05,319 And after a long search, they 157 00:06:05,320 --> 00:06:07,749 found one here that was found 158 00:06:07,750 --> 00:06:10,179 in a hedge and it was 18 centimeters 159 00:06:10,180 --> 00:06:12,759 long. And it was dyed blond. 160 00:06:12,760 --> 00:06:14,919 And on a surveillance 161 00:06:14,920 --> 00:06:17,229 video of a tram, they 162 00:06:17,230 --> 00:06:19,389 could find a person that 163 00:06:19,390 --> 00:06:22,629 fit this profile of this hair style, 164 00:06:22,630 --> 00:06:24,159 which led to an arrest. 165 00:06:24,160 --> 00:06:26,229 This was a 17 year old 166 00:06:26,230 --> 00:06:27,829 Afghan refugee. 167 00:06:27,830 --> 00:06:30,399 And he confessed 168 00:06:30,400 --> 00:06:32,619 that he committed the crime. 169 00:06:32,620 --> 00:06:34,569 And afterwards, the police said that 170 00:06:34,570 --> 00:06:37,419 would, with extended DNA analysis, 171 00:06:37,420 --> 00:06:39,279 the murderer could have been found much 172 00:06:39,280 --> 00:06:40,629 sooner. 173 00:06:40,630 --> 00:06:42,849 But what is this extended DNA 174 00:06:42,850 --> 00:06:43,929 analysis? 175 00:06:43,930 --> 00:06:46,289 There is now a new law 176 00:06:46,290 --> 00:06:48,909 to come in yet 177 00:06:48,910 --> 00:06:50,289 to come on. 178 00:06:50,290 --> 00:06:52,629 Not already here yet, but proposed 179 00:06:52,630 --> 00:06:54,909 by Baaden Vuitton Bag, which is now 180 00:06:54,910 --> 00:06:57,069 governed by the Green Party and 181 00:06:57,070 --> 00:06:59,229 Bavaria, which is governed by 182 00:06:59,230 --> 00:07:00,639 the Conservative Party. 183 00:07:00,640 --> 00:07:02,319 See as you. 184 00:07:02,320 --> 00:07:04,599 And they wish to draw more conclusions 185 00:07:04,600 --> 00:07:06,039 from our DNA. 186 00:07:06,040 --> 00:07:07,959 So they want to look at the eye color, 187 00:07:07,960 --> 00:07:10,119 the hair color, the skin color 188 00:07:10,120 --> 00:07:12,399 and the biological age. 189 00:07:12,400 --> 00:07:14,499 And the Conservative Party 190 00:07:14,500 --> 00:07:16,899 also wants to look at the biogeographical 191 00:07:16,900 --> 00:07:19,269 ancestry that can be determined 192 00:07:19,270 --> 00:07:20,739 from the DNA. 193 00:07:20,740 --> 00:07:23,319 And this is basically just 194 00:07:23,320 --> 00:07:25,659 an synonym for ethnicity. 195 00:07:25,660 --> 00:07:28,809 And some may even say for race 196 00:07:28,810 --> 00:07:29,810 and 197 00:07:31,330 --> 00:07:33,489 the law for extended DNA 198 00:07:33,490 --> 00:07:35,769 analysis is often referred to as 199 00:07:35,770 --> 00:07:38,379 the DNA facial composite. 200 00:07:38,380 --> 00:07:39,380 And 201 00:07:40,960 --> 00:07:43,299 however, in science terms, it's called 202 00:07:43,300 --> 00:07:45,969 forensic DNA, pheno typing. 203 00:07:45,970 --> 00:07:48,309 And this 204 00:07:48,310 --> 00:07:50,529 means that you look at genes 205 00:07:50,530 --> 00:07:52,719 which could determine the looks or 206 00:07:52,720 --> 00:07:55,180 other characteristics of a person 207 00:07:56,620 --> 00:07:58,869 just from the DNA and 208 00:07:58,870 --> 00:07:59,859 the genes. 209 00:07:59,860 --> 00:08:01,959 So what do you do hear is 210 00:08:01,960 --> 00:08:04,089 that you check for single nucleotide 211 00:08:04,090 --> 00:08:06,399 polymorphisms or call 212 00:08:06,400 --> 00:08:08,169 also called snips. 213 00:08:08,170 --> 00:08:10,239 And these snips are basically just 214 00:08:10,240 --> 00:08:12,519 single based mutations that you can see 215 00:08:12,520 --> 00:08:14,149 here. The A and the C. 216 00:08:15,640 --> 00:08:18,609 They can be everywhere in the DNA 217 00:08:18,610 --> 00:08:20,739 and with genome 218 00:08:20,740 --> 00:08:23,019 wide association studies or G 219 00:08:23,020 --> 00:08:25,269 was you can associate 220 00:08:25,270 --> 00:08:27,159 snips with the phenotype. 221 00:08:27,160 --> 00:08:29,439 So some people who have a C here 222 00:08:29,440 --> 00:08:31,299 would look different than people who have 223 00:08:31,300 --> 00:08:34,298 an A or have a different trait. 224 00:08:34,299 --> 00:08:36,339 These snips, however, do not have to be 225 00:08:36,340 --> 00:08:36,849 causal. 226 00:08:36,850 --> 00:08:39,038 They can just be a correlation, 227 00:08:39,039 --> 00:08:40,509 a statistical correlation. 228 00:08:42,220 --> 00:08:44,379 This looks like this. 229 00:08:44,380 --> 00:08:46,659 In our graphic here, 230 00:08:46,660 --> 00:08:48,909 if you have, for example, the same 231 00:08:48,910 --> 00:08:51,069 snip here for each person, 232 00:08:51,070 --> 00:08:53,529 but a different one for each person 233 00:08:53,530 --> 00:08:56,169 here or for these persons, 234 00:08:56,170 --> 00:08:58,239 the two first two are the same and the 235 00:08:58,240 --> 00:08:59,799 third one differs. 236 00:08:59,800 --> 00:09:01,869 But in the genes, you could maybe 237 00:09:01,870 --> 00:09:04,389 determine from the snips, 238 00:09:04,390 --> 00:09:06,819 the blond hair of this person, 239 00:09:06,820 --> 00:09:08,939 the brown hair of the second person 240 00:09:08,940 --> 00:09:11,109 and the red hair of the third 241 00:09:11,110 --> 00:09:13,179 person or 242 00:09:13,180 --> 00:09:15,729 the blue eyes of the first person, 243 00:09:15,730 --> 00:09:18,519 the brown eyes of the second person 244 00:09:18,520 --> 00:09:20,649 and the green eyes of the third person. 245 00:09:22,210 --> 00:09:24,369 Snip analysis is basically just done 246 00:09:24,370 --> 00:09:26,559 by checking 247 00:09:26,560 --> 00:09:28,329 for a known sequence. 248 00:09:28,330 --> 00:09:30,429 And then you have label bases 249 00:09:30,430 --> 00:09:32,529 and you can check which base binds 250 00:09:32,530 --> 00:09:33,530 here. 251 00:09:34,000 --> 00:09:35,859 This can be done by next generation 252 00:09:35,860 --> 00:09:37,929 sequencing, which is a quick form 253 00:09:37,930 --> 00:09:40,359 of sequencing DNA 254 00:09:40,360 --> 00:09:41,769 and a very novel technique. 255 00:09:43,570 --> 00:09:45,849 And you can do multiple snip 256 00:09:45,850 --> 00:09:48,039 analysis so you can check for 257 00:09:48,040 --> 00:09:49,899 several snips at once. 258 00:09:49,900 --> 00:09:52,029 So it's quite fast and 259 00:09:52,030 --> 00:09:53,559 it just gives you a statistical 260 00:09:53,560 --> 00:09:55,689 correlation of which person has 261 00:09:55,690 --> 00:09:56,919 which snips 262 00:09:58,210 --> 00:10:00,909 and and looks like 263 00:10:00,910 --> 00:10:03,099 what. And soon 264 00:10:03,100 --> 00:10:05,169 maybe we could also do a whole genome 265 00:10:05,170 --> 00:10:07,659 sequencing and then check for 266 00:10:07,660 --> 00:10:10,000 the bases that are mutated. 267 00:10:12,010 --> 00:10:14,139 The accuracy, if you 268 00:10:14,140 --> 00:10:16,569 check for the biogeographic ancestry, 269 00:10:16,570 --> 00:10:19,719 if you just go for Continental ancestry, 270 00:10:19,720 --> 00:10:22,089 this is about 99 percent. 271 00:10:22,090 --> 00:10:24,489 So you can determine 272 00:10:24,490 --> 00:10:27,129 if you're European or African 273 00:10:27,130 --> 00:10:29,529 or Native American, for example. 274 00:10:29,530 --> 00:10:31,959 But what you get sometimes 275 00:10:31,960 --> 00:10:34,209 from the media. Twenty percent Spanish 276 00:10:34,210 --> 00:10:36,279 or 20 percent Turkish. 277 00:10:36,280 --> 00:10:37,719 Or something that doesn't work. 278 00:10:37,720 --> 00:10:39,969 That's too complicated. 279 00:10:39,970 --> 00:10:42,399 However, hair color can be determined 280 00:10:42,400 --> 00:10:44,799 with an accuracy between 70 to 90 281 00:10:44,800 --> 00:10:47,559 percent. So in 70 to 90 percent, 282 00:10:47,560 --> 00:10:49,959 the result will be correct 283 00:10:51,700 --> 00:10:53,829 eye color. The result for blue or 284 00:10:53,830 --> 00:10:55,899 brown hair will be correct in 285 00:10:55,900 --> 00:10:57,999 over 90 percent of the cases. 286 00:10:58,000 --> 00:11:00,249 Intermediate forms will be corrected. 287 00:11:00,250 --> 00:11:02,919 About 70 percent of the cases 288 00:11:02,920 --> 00:11:04,240 and skin color 289 00:11:05,350 --> 00:11:07,419 will be correct in 290 00:11:07,420 --> 00:11:09,549 about 80 to 90 percent of 291 00:11:09,550 --> 00:11:11,829 the cases to check 292 00:11:11,830 --> 00:11:12,939 for this more closely. 293 00:11:12,940 --> 00:11:14,859 Here are some examples from a paper by 294 00:11:14,860 --> 00:11:17,139 Kaiser. I and 295 00:11:17,140 --> 00:11:19,299 you can see here that for the brown 296 00:11:19,300 --> 00:11:21,699 hair, this was not quite 297 00:11:21,700 --> 00:11:24,279 as sure. So it was something between 298 00:11:24,280 --> 00:11:25,540 brown and black. 299 00:11:26,740 --> 00:11:28,929 The dark shade of the skin, however, 300 00:11:28,930 --> 00:11:30,989 was determined correctly 301 00:11:30,990 --> 00:11:33,969 at about 80 percent certainty. 302 00:11:33,970 --> 00:11:36,579 And the brown hair, brown eyes 303 00:11:36,580 --> 00:11:39,009 were also correct with over 90 percent 304 00:11:39,010 --> 00:11:40,010 certainty. 305 00:11:41,410 --> 00:11:43,659 For the brown hair person here, 306 00:11:43,660 --> 00:11:45,819 you have about 66 percent 307 00:11:45,820 --> 00:11:47,889 likelihood for brown hair. 308 00:11:47,890 --> 00:11:50,769 And also, 309 00:11:50,770 --> 00:11:52,959 it could have gone even black 310 00:11:52,960 --> 00:11:55,029 or blond here in 311 00:11:55,030 --> 00:11:57,519 the shade here was not quite as accurate. 312 00:11:57,520 --> 00:11:59,499 It was 50/50. 313 00:11:59,500 --> 00:12:01,839 And for brown 314 00:12:01,840 --> 00:12:04,509 hair, brown eyes, it was 72 315 00:12:04,510 --> 00:12:07,579 percent likelihood of accuracy. 316 00:12:07,580 --> 00:12:09,669 And this 317 00:12:09,670 --> 00:12:11,859 blond person here had 318 00:12:11,860 --> 00:12:14,109 about 62 percent accuracy 319 00:12:14,110 --> 00:12:16,029 of the blond hair. 320 00:12:16,030 --> 00:12:18,159 97 percent of accuracy 321 00:12:18,160 --> 00:12:20,199 for the light shade of the skin. 322 00:12:20,200 --> 00:12:22,779 And also 95 percent accuracy 323 00:12:22,780 --> 00:12:24,669 of the blue eyes. 324 00:12:24,670 --> 00:12:26,979 So this was quite 325 00:12:26,980 --> 00:12:29,019 accurate, this prediction. 326 00:12:29,020 --> 00:12:32,019 And here, the red haired person 327 00:12:32,020 --> 00:12:34,299 of 92 percent red hair, 328 00:12:34,300 --> 00:12:36,129 however, was still skin tone. 329 00:12:36,130 --> 00:12:38,319 Again, this was quite undecided. 330 00:12:38,320 --> 00:12:40,329 So it was, again, 50/50. 331 00:12:40,330 --> 00:12:42,909 So it's dangerous to draw a conclusion 332 00:12:42,910 --> 00:12:45,339 here just because one is 333 00:12:45,340 --> 00:12:47,109 higher than the other. 334 00:12:47,110 --> 00:12:49,449 And also for the brown hair, brown 335 00:12:49,450 --> 00:12:51,519 eyes, there was just 46 percent 336 00:12:51,520 --> 00:12:52,520 certainty here. 337 00:12:54,320 --> 00:12:56,599 Then you can check for the biological 338 00:12:56,600 --> 00:12:59,209 age, which you can do doing 339 00:12:59,210 --> 00:13:00,949 checking for the DNA methylation 340 00:13:00,950 --> 00:13:01,909 patterns. 341 00:13:01,910 --> 00:13:04,249 So our DNA is modified by methyl 342 00:13:04,250 --> 00:13:06,319 groups and this can 343 00:13:06,320 --> 00:13:08,449 be done to so-called age specific 344 00:13:08,450 --> 00:13:10,099 CPG islands. 345 00:13:10,100 --> 00:13:12,319 So CPG islands are just a quick 346 00:13:12,320 --> 00:13:14,509 repetition between of 347 00:13:14,510 --> 00:13:16,609 C and G before 348 00:13:16,610 --> 00:13:17,629 a gene. 349 00:13:17,630 --> 00:13:19,639 So in front of a gene and these can 350 00:13:19,640 --> 00:13:22,189 regulate the expression of a gene. 351 00:13:22,190 --> 00:13:24,469 So if a gene 352 00:13:24,470 --> 00:13:27,889 is later brought up in your looks 353 00:13:27,890 --> 00:13:30,799 and there's some 354 00:13:30,800 --> 00:13:32,389 kind of looks like this that you have, 355 00:13:32,390 --> 00:13:34,459 for example, an older person on the 356 00:13:34,460 --> 00:13:36,559 top who has 357 00:13:36,560 --> 00:13:39,109 has methylated 358 00:13:39,110 --> 00:13:41,599 CBG islands in front of the genes. 359 00:13:41,600 --> 00:13:43,939 However, a younger 360 00:13:43,940 --> 00:13:46,129 person in the middle who doesn't have age 361 00:13:46,130 --> 00:13:48,469 specific CBG 362 00:13:48,470 --> 00:13:50,569 methylation and an 363 00:13:50,570 --> 00:13:52,999 intermediate age for a person 364 00:13:53,000 --> 00:13:55,880 with just this one methylation. 365 00:13:57,290 --> 00:13:59,749 You can analyze this during 366 00:13:59,750 --> 00:14:02,389 doing a chemical treatment of the DNA. 367 00:14:02,390 --> 00:14:04,489 So your DNA, which has 368 00:14:04,490 --> 00:14:06,559 been methylated at the site or seen 369 00:14:08,180 --> 00:14:10,369 that is treated. 370 00:14:10,370 --> 00:14:12,559 And then you get a change of 371 00:14:12,560 --> 00:14:14,149 base. If it was an methylated. 372 00:14:14,150 --> 00:14:16,279 So you can see in the lower part it 373 00:14:16,280 --> 00:14:18,409 has changed to a different 374 00:14:18,410 --> 00:14:20,059 base called Uracil. 375 00:14:20,060 --> 00:14:22,069 And then you can sequence the DNA and see 376 00:14:22,070 --> 00:14:23,539 how much it was exchanged. 377 00:14:25,160 --> 00:14:27,229 The accuracy here, there is a 378 00:14:27,230 --> 00:14:30,349 mean arrow, about three to five years. 379 00:14:30,350 --> 00:14:32,449 But some results will have up to 380 00:14:32,450 --> 00:14:34,339 20 years. 381 00:14:34,340 --> 00:14:35,450 Errors and 382 00:14:36,590 --> 00:14:39,049 there are some diseases 383 00:14:39,050 --> 00:14:40,849 that can confound the results. 384 00:14:40,850 --> 00:14:43,759 For example, cancer or anemia. 385 00:14:43,760 --> 00:14:45,919 And also, it's a problem because 386 00:14:45,920 --> 00:14:48,379 this is tissue specific methylation. 387 00:14:48,380 --> 00:14:50,419 If you have mixed tissues. 388 00:14:50,420 --> 00:14:52,369 So usually this is done from blood and 389 00:14:52,370 --> 00:14:53,539 saliva. 390 00:14:53,540 --> 00:14:55,879 And if you have a mixture of other 391 00:14:55,880 --> 00:14:58,069 tissues, then this can be 392 00:14:58,070 --> 00:14:59,750 more accurate. 393 00:15:00,950 --> 00:15:02,839 So basically, what's advertised, if you 394 00:15:02,840 --> 00:15:05,149 hear the term DNA facial 395 00:15:05,150 --> 00:15:07,369 composite, is this facial 396 00:15:07,370 --> 00:15:08,869 composite here. 397 00:15:08,870 --> 00:15:11,119 But so this is probably the first 398 00:15:11,120 --> 00:15:12,859 thing that people think. 399 00:15:12,860 --> 00:15:14,959 However, the reality 400 00:15:14,960 --> 00:15:16,820 is more something like this. 401 00:15:18,920 --> 00:15:20,989 And maybe you have a little bit 402 00:15:20,990 --> 00:15:23,299 the age of these persons. 403 00:15:23,300 --> 00:15:26,359 And then only maybe. 404 00:15:26,360 --> 00:15:28,339 And I'd like to now for you to look 405 00:15:28,340 --> 00:15:30,769 around you and see how many people 406 00:15:30,770 --> 00:15:33,289 fit either of these profiles. 407 00:15:33,290 --> 00:15:35,329 And just to determine this for you for a 408 00:15:35,330 --> 00:15:36,330 second. 409 00:15:37,610 --> 00:15:39,739 And so this is 410 00:15:39,740 --> 00:15:42,199 just an approximation 411 00:15:42,200 --> 00:15:43,609 of the phenotype. 412 00:15:44,810 --> 00:15:47,359 And as you might remember 413 00:15:47,360 --> 00:15:49,819 or know, even the DNA 414 00:15:49,820 --> 00:15:52,699 does not have to equal the phenotype. 415 00:15:52,700 --> 00:15:55,369 So people who have black hair 416 00:15:55,370 --> 00:15:57,649 in young years might have gray 417 00:15:57,650 --> 00:15:59,989 hair. Now, some 418 00:15:59,990 --> 00:16:02,539 people are born black. 419 00:16:02,540 --> 00:16:05,629 An hour later of a different skin color. 420 00:16:05,630 --> 00:16:07,849 And some people 421 00:16:07,850 --> 00:16:11,059 may well have X Y chromosomes, 422 00:16:11,060 --> 00:16:12,889 but are female. 423 00:16:12,890 --> 00:16:15,079 And other factors 424 00:16:15,080 --> 00:16:16,369 may also be. 425 00:16:16,370 --> 00:16:18,499 For example, kids who have 426 00:16:18,500 --> 00:16:20,629 blond hair and grow up to be dark 427 00:16:20,630 --> 00:16:21,649 haired adults. 428 00:16:22,760 --> 00:16:24,979 Also, hair color is 429 00:16:24,980 --> 00:16:27,259 always very, very easy to change. 430 00:16:27,260 --> 00:16:29,239 And purple is not in our genes. 431 00:16:31,220 --> 00:16:33,889 And also mixed 432 00:16:33,890 --> 00:16:34,909 ancestry. 433 00:16:34,910 --> 00:16:37,069 So having ancestors from different 434 00:16:37,070 --> 00:16:39,229 continents of the world is also 435 00:16:39,230 --> 00:16:41,779 a problem, often with determining 436 00:16:41,780 --> 00:16:42,780 the ancestry. 437 00:16:44,210 --> 00:16:46,519 But they are more problems with forensic 438 00:16:46,520 --> 00:16:48,289 DNA, pheno typing. 439 00:16:48,290 --> 00:16:51,049 And the first thing and most important 440 00:16:51,050 --> 00:16:53,359 ethical problem is the suspicion 441 00:16:53,360 --> 00:16:54,589 reversal. 442 00:16:54,590 --> 00:16:56,929 So the police usually 443 00:16:56,930 --> 00:16:59,449 has to prove that you are guilty. 444 00:16:59,450 --> 00:17:01,579 But if you're a suspect, just due to 445 00:17:01,580 --> 00:17:03,949 your looks, then with 446 00:17:03,950 --> 00:17:06,019 the extended DNA analysis, you will 447 00:17:06,020 --> 00:17:08,209 be a non suspect person of 448 00:17:08,210 --> 00:17:09,299 interest. 449 00:17:09,300 --> 00:17:11,568 And so 450 00:17:11,569 --> 00:17:13,969 you will be kind of connected 451 00:17:13,970 --> 00:17:15,318 to the crime. 452 00:17:15,319 --> 00:17:17,509 We are to the probe 453 00:17:17,510 --> 00:17:19,999 without having done anything. 454 00:17:21,109 --> 00:17:23,179 And now you have to prove that you're 455 00:17:23,180 --> 00:17:24,439 innocent. 456 00:17:24,440 --> 00:17:26,749 And Victor Taum, social 457 00:17:26,750 --> 00:17:29,029 scientists from Frankfurt, he said 458 00:17:29,030 --> 00:17:31,129 suspicionless seizure of persons 459 00:17:31,130 --> 00:17:33,349 and then demanding proof of that person's 460 00:17:33,350 --> 00:17:35,959 innocence inverts our structural 461 00:17:35,960 --> 00:17:38,149 arrangement of power between 462 00:17:38,150 --> 00:17:40,609 law enforcement and the individual. 463 00:17:40,610 --> 00:17:42,889 So this is one of the 464 00:17:42,890 --> 00:17:45,260 most important things that you see here. 465 00:17:46,940 --> 00:17:49,400 With forensic DNA, pheno typing 466 00:17:51,020 --> 00:17:53,149 and oftentimes 467 00:17:53,150 --> 00:17:54,319 signs. 468 00:17:54,320 --> 00:17:56,449 Is supposed to equal 469 00:17:56,450 --> 00:17:58,549 facts and DNA is 470 00:17:58,550 --> 00:18:00,709 an unbiased witness because 471 00:18:00,710 --> 00:18:02,869 it's in Farlam Bill and factual 472 00:18:02,870 --> 00:18:04,969 because it's based on science and 473 00:18:04,970 --> 00:18:07,129 it's done by scientists. 474 00:18:07,130 --> 00:18:09,289 However, you have 475 00:18:09,290 --> 00:18:12,019 only an estimation about how 476 00:18:12,020 --> 00:18:14,899 people could possibly look like. 477 00:18:14,900 --> 00:18:17,089 And therefore, it has to be 478 00:18:17,090 --> 00:18:19,599 interpreted by a human being. 479 00:18:19,600 --> 00:18:21,679 And therefore, it's prone 480 00:18:21,680 --> 00:18:23,119 to bias. 481 00:18:23,120 --> 00:18:25,549 And one of the most important biases 482 00:18:25,550 --> 00:18:28,099 here is the confirmation bias. 483 00:18:28,100 --> 00:18:30,329 And I have to say that 484 00:18:30,330 --> 00:18:32,589 everyone and each of us, however, 485 00:18:32,590 --> 00:18:35,179 are aware. We feel we are 486 00:18:35,180 --> 00:18:37,019 we are always biased. 487 00:18:37,020 --> 00:18:39,109 That's how our brain works. 488 00:18:39,110 --> 00:18:41,179 And our brain likes to think more 489 00:18:41,180 --> 00:18:43,399 positively and 490 00:18:43,400 --> 00:18:45,859 not so much negatively. 491 00:18:45,860 --> 00:18:48,679 So if we anticipate an outcome, 492 00:18:48,680 --> 00:18:50,779 we are more likely to prove that this 493 00:18:50,780 --> 00:18:52,579 outcome is true. 494 00:18:52,580 --> 00:18:55,069 If we have, we pay selective 495 00:18:55,070 --> 00:18:57,409 attention to items of interest 496 00:18:57,410 --> 00:18:59,329 and disregard the contradictory 497 00:18:59,330 --> 00:19:02,359 information just unwillingly. 498 00:19:02,360 --> 00:19:04,789 And we will have a positivity bias. 499 00:19:04,790 --> 00:19:06,979 So we like to confirm what we think is 500 00:19:06,980 --> 00:19:08,059 true. 501 00:19:08,060 --> 00:19:10,099 And also, if we have heard something 502 00:19:10,100 --> 00:19:12,769 before, we think it's more true. 503 00:19:12,770 --> 00:19:13,849 It doesn't make sense. 504 00:19:13,850 --> 00:19:16,369 But it's how our brain works. 505 00:19:16,370 --> 00:19:18,769 But therefore, DNA is not 506 00:19:18,770 --> 00:19:19,999 an unbiased witness. 507 00:19:21,200 --> 00:19:23,659 And also, there's a question 508 00:19:23,660 --> 00:19:25,309 of privacy here. 509 00:19:25,310 --> 00:19:27,679 So the last I posit the hypothesis 510 00:19:27,680 --> 00:19:29,929 is that openly visible features 511 00:19:29,930 --> 00:19:31,999 determined from the DNA are not 512 00:19:32,000 --> 00:19:33,449 private. 513 00:19:33,450 --> 00:19:36,019 And however, here, the appearance 514 00:19:36,020 --> 00:19:38,809 does not equal the statistical 515 00:19:38,810 --> 00:19:40,039 correlation. 516 00:19:40,040 --> 00:19:42,649 And also, if you have a real witness 517 00:19:42,650 --> 00:19:44,869 who saw your features, 518 00:19:44,870 --> 00:19:47,059 you have more context and 519 00:19:47,060 --> 00:19:49,159 you can say, for example, 520 00:19:49,160 --> 00:19:51,139 the person was fleeing the scene or 521 00:19:51,140 --> 00:19:52,140 something. 522 00:19:53,810 --> 00:19:56,059 Gene's on the looks and biogeographic 523 00:19:56,060 --> 00:19:58,669 background will always tell us more 524 00:19:58,670 --> 00:20:00,289 than we ask. 525 00:20:00,290 --> 00:20:02,719 And most important maybe 526 00:20:02,720 --> 00:20:03,949 is skin color. 527 00:20:03,950 --> 00:20:06,589 So fair skin is much more likely 528 00:20:06,590 --> 00:20:09,169 to get skin cancer or 529 00:20:09,170 --> 00:20:11,359 darker skin. People have a higher risk 530 00:20:11,360 --> 00:20:12,859 for heart disease. 531 00:20:12,860 --> 00:20:14,959 And also 532 00:20:14,960 --> 00:20:17,029 age is a factor 533 00:20:17,030 --> 00:20:18,339 for many diseases. 534 00:20:22,510 --> 00:20:24,579 Then I'd like to come to 535 00:20:24,580 --> 00:20:26,709 the point of discrimination. 536 00:20:26,710 --> 00:20:29,049 So by definition, forensic 537 00:20:29,050 --> 00:20:31,359 DNA, pheno typing discriminates 538 00:20:31,360 --> 00:20:34,149 people with different ethnicities. 539 00:20:34,150 --> 00:20:36,459 So if you think 540 00:20:36,460 --> 00:20:38,889 about it, the relevant suspect group 541 00:20:38,890 --> 00:20:41,109 will always be the minority 542 00:20:41,110 --> 00:20:43,299 of the people who actually live in 543 00:20:43,300 --> 00:20:44,469 this area. 544 00:20:44,470 --> 00:20:47,049 Because if you find the looks 545 00:20:47,050 --> 00:20:49,359 of the predominant population, 546 00:20:49,360 --> 00:20:51,519 this will be too large of a suspect 547 00:20:51,520 --> 00:20:53,889 group and 548 00:20:53,890 --> 00:20:55,659 openly discussed forensic DNA. 549 00:20:55,660 --> 00:20:57,969 Phenotypes of a DNA trace 550 00:20:57,970 --> 00:21:00,579 to a crime will probably 551 00:21:00,580 --> 00:21:02,649 lead to more hate crimes 552 00:21:02,650 --> 00:21:04,149 or add to the hate crimes that are 553 00:21:04,150 --> 00:21:07,299 already in place here. 554 00:21:07,300 --> 00:21:09,369 And so if deployed forensic 555 00:21:09,370 --> 00:21:10,719 DNA phenotype, it needs to be 556 00:21:10,720 --> 00:21:13,029 confidential and needs to avoid 557 00:21:13,030 --> 00:21:15,700 stigmatization of whole subpopulations. 558 00:21:17,760 --> 00:21:20,039 And therefore, the 559 00:21:20,040 --> 00:21:22,199 debate here is always very 560 00:21:22,200 --> 00:21:24,389 emotional because the forensic DNA 561 00:21:24,390 --> 00:21:26,639 pheno typing law was proposed 562 00:21:26,640 --> 00:21:29,009 after a very brutal rape case. 563 00:21:29,010 --> 00:21:31,109 And the first people who demanded 564 00:21:31,110 --> 00:21:33,299 forensic DNA, pheno typing were actually 565 00:21:33,300 --> 00:21:35,729 right, extremism groups. 566 00:21:35,730 --> 00:21:37,829 And the 567 00:21:37,830 --> 00:21:40,889 new law probably would have not helped 568 00:21:40,890 --> 00:21:43,529 the investigation because 569 00:21:43,530 --> 00:21:46,649 the perpetrator had strikingly 570 00:21:46,650 --> 00:21:47,909 dyed hair. 571 00:21:47,910 --> 00:21:50,249 And that was the reason he was identified 572 00:21:50,250 --> 00:21:51,420 on the surveillance video. 573 00:21:54,250 --> 00:21:56,319 The outlook, actually, for 574 00:21:56,320 --> 00:21:58,719 forensic DNA, pheno typing, it's 575 00:21:58,720 --> 00:22:00,489 getting more accurate. 576 00:22:00,490 --> 00:22:02,529 So people are trying now to determine the 577 00:22:02,530 --> 00:22:04,989 body height, the body stature, 578 00:22:04,990 --> 00:22:05,889 hair loss. 579 00:22:05,890 --> 00:22:08,049 So if you're a bald person, the 580 00:22:08,050 --> 00:22:10,269 hair structure and also the face 581 00:22:10,270 --> 00:22:12,340 structure, and you can actually see 582 00:22:13,440 --> 00:22:15,609 on the image approximations 583 00:22:15,610 --> 00:22:18,129 of face structure based on 584 00:22:18,130 --> 00:22:19,390 snip analysis. 585 00:22:21,930 --> 00:22:24,029 So extend a DNA analysis, 586 00:22:24,030 --> 00:22:26,339 whole uses the whole genome 587 00:22:26,340 --> 00:22:28,529 wide association studies 588 00:22:28,530 --> 00:22:30,919 to approximate looks. 589 00:22:30,920 --> 00:22:34,079 H biogeographical ancestry 590 00:22:34,080 --> 00:22:37,079 and maybe soon much more. 591 00:22:37,080 --> 00:22:39,659 And it is not a means 592 00:22:39,660 --> 00:22:41,789 of identification, but 593 00:22:41,790 --> 00:22:43,649 approximation. 594 00:22:43,650 --> 00:22:46,739 And by definition, it's discriminatory 595 00:22:46,740 --> 00:22:49,469 and it will target minorities. 596 00:22:49,470 --> 00:22:51,899 And since it is biased, 597 00:22:51,900 --> 00:22:53,909 it can always lead into the wrong 598 00:22:53,910 --> 00:22:54,910 direction. 599 00:22:56,440 --> 00:22:58,269 So what can be done here? 600 00:22:59,370 --> 00:23:01,509 Forensic forensic DNA, pheno typing 601 00:23:01,510 --> 00:23:02,510 is to come. 602 00:23:03,560 --> 00:23:05,619 There's only as a method 603 00:23:05,620 --> 00:23:08,139 of F last resort when everything 604 00:23:08,140 --> 00:23:09,549 else has failed. 605 00:23:09,550 --> 00:23:11,559 It has to be confidential. 606 00:23:11,560 --> 00:23:13,689 So there are no hate crimes due to this 607 00:23:13,690 --> 00:23:15,219 investigation. 608 00:23:15,220 --> 00:23:17,349 And precautions have to be done 609 00:23:17,350 --> 00:23:19,210 to avoid confirmation bias 610 00:23:20,560 --> 00:23:22,929 because no one and nothing 611 00:23:22,930 --> 00:23:24,999 is infallible and no 612 00:23:25,000 --> 00:23:27,009 one is unbiased. 613 00:23:27,010 --> 00:23:30,279 And every wrongful investigation 614 00:23:30,280 --> 00:23:33,579 of a suspect leaves the real perpetrator 615 00:23:33,580 --> 00:23:34,580 unprosecuted. 616 00:23:36,670 --> 00:23:38,739 So back to the 617 00:23:38,740 --> 00:23:39,969 new law. 618 00:23:39,970 --> 00:23:42,189 So the Conference 619 00:23:42,190 --> 00:23:44,409 of Ministers in 620 00:23:44,410 --> 00:23:46,209 of Internal Affairs in Minister 621 00:23:46,210 --> 00:23:48,339 Conference supports this law 622 00:23:48,340 --> 00:23:51,099 inside of the security package 623 00:23:51,100 --> 00:23:53,949 and habit oil, the interior ministers 624 00:23:53,950 --> 00:23:56,439 of North Irma's failure 625 00:23:56,440 --> 00:23:58,779 in Ovie. He said 626 00:23:58,780 --> 00:24:01,449 nobody understands why forensic 627 00:24:01,450 --> 00:24:03,549 scientists can form a very 628 00:24:03,550 --> 00:24:06,459 precise facial composite 629 00:24:06,460 --> 00:24:08,559 from very small traces. 630 00:24:08,560 --> 00:24:10,629 But our police cannot use 631 00:24:10,630 --> 00:24:11,630 them. 632 00:24:12,430 --> 00:24:14,890 Well, maybe now you understand. 633 00:24:18,040 --> 00:24:20,199 So I'd like to thank you very 634 00:24:20,200 --> 00:24:21,909 much for your attention. 635 00:24:21,910 --> 00:24:24,219 And if you'd like to know more about 636 00:24:24,220 --> 00:24:26,859 me, you can find me on my Web site. 637 00:24:26,860 --> 00:24:28,779 And if you want to know more about 638 00:24:28,780 --> 00:24:30,999 forensic DNA, pheno typing and 639 00:24:31,000 --> 00:24:33,519 ethical issues with this professor, 640 00:24:33,520 --> 00:24:35,679 the only car and on our Leaphart are 641 00:24:35,680 --> 00:24:37,899 doing very great work on this 642 00:24:37,900 --> 00:24:40,029 at the University of Vyborg. 643 00:24:40,030 --> 00:24:41,889 And this is their Web site. 644 00:24:41,890 --> 00:24:44,200 And I'm now open for questions. 645 00:24:54,070 --> 00:24:56,019 So thank you very much for an interesting 646 00:24:56,020 --> 00:24:57,309 talk, if you have any questions. 647 00:24:57,310 --> 00:24:58,689 There are two microphones on each of the 648 00:24:58,690 --> 00:25:00,459 aisles. And the first question from 649 00:25:00,460 --> 00:25:01,460 microphone number four. 650 00:25:03,070 --> 00:25:05,559 I wanted to ask how clean eventually 651 00:25:05,560 --> 00:25:07,749 the DNA sample should be 652 00:25:07,750 --> 00:25:10,209 to proceed with this procedure. 653 00:25:10,210 --> 00:25:12,579 It probably has to be very, very clean 654 00:25:12,580 --> 00:25:14,709 because I'm not quite sure 655 00:25:14,710 --> 00:25:17,039 how easy you can determine 656 00:25:17,040 --> 00:25:18,219 distinguished them. 657 00:25:18,220 --> 00:25:20,589 I think it should be completely clean. 658 00:25:20,590 --> 00:25:22,899 So if you have maybe DNA from 659 00:25:22,900 --> 00:25:25,149 a rape, which is mingled with 660 00:25:25,150 --> 00:25:27,579 the victim's DNA, it's difficult 661 00:25:27,580 --> 00:25:29,469 to get this out. 662 00:25:29,470 --> 00:25:30,470 However, if you can 663 00:25:31,840 --> 00:25:34,149 control it with comparing 664 00:25:34,150 --> 00:25:36,279 it with the DNA of the victim, 665 00:25:36,280 --> 00:25:38,499 then maybe it can be done 666 00:25:38,500 --> 00:25:39,790 a little bit more accurately. 667 00:25:41,120 --> 00:25:43,039 Quester from the Internet. 668 00:25:43,040 --> 00:25:44,959 Yep, the Internet wants to know what's 669 00:25:44,960 --> 00:25:47,429 your opinion about using DNA 670 00:25:47,430 --> 00:25:49,909 fingerprint of someone 671 00:25:49,910 --> 00:25:50,899 else? 672 00:25:50,900 --> 00:25:53,149 Probably key to identify persons 673 00:25:53,150 --> 00:25:55,639 on a distributed ledger technology 674 00:25:55,640 --> 00:25:56,640 network. 675 00:26:00,840 --> 00:26:02,969 I don't know what a distributed 676 00:26:02,970 --> 00:26:05,189 ledger technology network 677 00:26:05,190 --> 00:26:06,139 is. 678 00:26:06,140 --> 00:26:07,559 Excuse me. 679 00:26:07,560 --> 00:26:09,659 Maybe you can ask this 680 00:26:09,660 --> 00:26:11,399 question on my blog. 681 00:26:11,400 --> 00:26:13,379 Just sent me an e-mail and we can discuss 682 00:26:13,380 --> 00:26:15,689 this, using it as 683 00:26:15,690 --> 00:26:17,819 identification for something I 684 00:26:17,820 --> 00:26:20,039 think right now is not really feasible, 685 00:26:20,040 --> 00:26:22,169 because to do it accurately, you need 686 00:26:22,170 --> 00:26:24,509 someone to to do it for 687 00:26:24,510 --> 00:26:26,609 you in a scientific lab under 688 00:26:26,610 --> 00:26:29,009 specific circumstances. 689 00:26:29,010 --> 00:26:31,319 And if you have DNA from someone 690 00:26:31,320 --> 00:26:33,509 else on your fingers, maybe you 691 00:26:33,510 --> 00:26:35,609 transfer this DNA and suddenly your 692 00:26:35,610 --> 00:26:38,059 ID doesn't work anymore. 693 00:26:38,060 --> 00:26:40,139 So this would 694 00:26:40,140 --> 00:26:42,899 be what I would be thinking about this 695 00:26:42,900 --> 00:26:43,900 microphone number two. 696 00:26:45,070 --> 00:26:47,639 Thank you for your very good talk. 697 00:26:47,640 --> 00:26:49,839 My question is, does 698 00:26:49,840 --> 00:26:51,749 seems to be a great difference between 699 00:26:51,750 --> 00:26:54,279 the opinion of science, how 700 00:26:54,280 --> 00:26:56,469 we should use this technology and 701 00:26:56,470 --> 00:26:57,470 politicians 702 00:26:58,810 --> 00:27:00,879 and this kind of conflict, I 703 00:27:00,880 --> 00:27:02,979 guess, could only be 704 00:27:02,980 --> 00:27:05,109 fought by the scientists 705 00:27:05,110 --> 00:27:06,699 against the politicians. 706 00:27:06,700 --> 00:27:08,919 Is a kind of movement of 707 00:27:08,920 --> 00:27:11,079 scientists who say, well, what's the 708 00:27:11,080 --> 00:27:13,119 politicians say? That we should use this. 709 00:27:13,120 --> 00:27:15,219 And it's so secure 710 00:27:15,220 --> 00:27:17,409 and not without 711 00:27:17,410 --> 00:27:19,619 any bias. Is there a movement 712 00:27:19,620 --> 00:27:21,759 to say they are talking bullshit? 713 00:27:21,760 --> 00:27:23,979 So this is basically what we only car and 714 00:27:23,980 --> 00:27:25,029 Lippard would be saying. 715 00:27:25,030 --> 00:27:26,289 They are not saying they are talking 716 00:27:26,290 --> 00:27:28,029 bullshit because they're raising 717 00:27:28,030 --> 00:27:30,189 awareness for the problems. 718 00:27:30,190 --> 00:27:32,260 And so I would never say 719 00:27:33,610 --> 00:27:35,739 something like that. But, um, 720 00:27:36,760 --> 00:27:39,789 if they are also scientists who propose 721 00:27:39,790 --> 00:27:42,009 to use this as 722 00:27:42,010 --> 00:27:44,469 as dear as help in forensic 723 00:27:44,470 --> 00:27:46,569 investigations, because 724 00:27:46,570 --> 00:27:49,069 a lot of us scientists, we are very 725 00:27:49,070 --> 00:27:51,219 well. We we don't look at 726 00:27:51,220 --> 00:27:53,199 the world as many people do. 727 00:27:53,200 --> 00:27:54,939 We think of it actually as well. 728 00:27:54,940 --> 00:27:56,679 This is just data and it's just an 729 00:27:56,680 --> 00:27:58,479 approximation. Of course, people know 730 00:27:58,480 --> 00:28:00,369 this as an approximation. 731 00:28:00,370 --> 00:28:02,439 And so we would be 732 00:28:02,440 --> 00:28:04,839 thinking, so if the DNA is of a blond 733 00:28:04,840 --> 00:28:05,949 haired person. 734 00:28:05,950 --> 00:28:07,899 Of course, that would just be paying more 735 00:28:07,900 --> 00:28:09,429 attention to a blond haired person. 736 00:28:09,430 --> 00:28:11,499 But if there is a brown hair person 737 00:28:11,500 --> 00:28:13,929 who is, uh, very much connected 738 00:28:13,930 --> 00:28:16,029 to the person in question, then you 739 00:28:16,030 --> 00:28:18,279 might even pay 740 00:28:18,280 --> 00:28:20,499 attention to this, even though the DNA 741 00:28:20,500 --> 00:28:22,659 analysis said something else, just 742 00:28:22,660 --> 00:28:24,999 because scientists are aware 743 00:28:25,000 --> 00:28:26,560 that the DNA can be wrong. 744 00:28:27,940 --> 00:28:29,889 I'm not saying that there are people who 745 00:28:29,890 --> 00:28:32,259 also understand that the DNA 746 00:28:32,260 --> 00:28:34,119 can lead us into the wrong direction, 747 00:28:34,120 --> 00:28:36,879 even in the police force and everywhere. 748 00:28:36,880 --> 00:28:39,849 But history has taught us that people 749 00:28:39,850 --> 00:28:42,159 a lot of times tend to go 750 00:28:42,160 --> 00:28:44,379 where the signs 751 00:28:44,380 --> 00:28:46,089 lead them, leads them here. 752 00:28:46,090 --> 00:28:48,339 And if the science has 753 00:28:48,340 --> 00:28:50,609 not been as accurate at 754 00:28:50,610 --> 00:28:52,900 that point, then maybe it doesn't. 755 00:28:54,070 --> 00:28:55,149 Yeah. 756 00:28:55,150 --> 00:28:57,579 It won't really work in this case. 757 00:28:59,330 --> 00:29:00,899 Number three. 758 00:29:00,900 --> 00:29:02,449 Thanks for your talk. 759 00:29:02,450 --> 00:29:04,549 And I just wondered if you 760 00:29:04,550 --> 00:29:06,619 create a profile on this 761 00:29:06,620 --> 00:29:08,729 that features a hair color 762 00:29:08,730 --> 00:29:10,849 or a skin color or eye color. 763 00:29:10,850 --> 00:29:12,079 And each of the features has a 764 00:29:12,080 --> 00:29:14,569 probability of something like point seven 765 00:29:14,570 --> 00:29:17,479 to be correct. And you multiply those 766 00:29:17,480 --> 00:29:19,669 probably to build his up, you will end 767 00:29:19,670 --> 00:29:21,739 up with a much lower probability 768 00:29:21,740 --> 00:29:24,049 for the whole profile to be actually 769 00:29:24,050 --> 00:29:26,119 right. Something like maybe 25 or 30 770 00:29:26,120 --> 00:29:27,469 percent or so. 771 00:29:27,470 --> 00:29:29,389 And so I wonder whether there are 772 00:29:29,390 --> 00:29:31,849 actually any ideas how this 773 00:29:31,850 --> 00:29:34,159 method should be used in practice 774 00:29:34,160 --> 00:29:35,419 for the police. 775 00:29:35,420 --> 00:29:38,819 And on the other hand, what 776 00:29:38,820 --> 00:29:41,089 value this kind of profiles 777 00:29:41,090 --> 00:29:43,069 should have in court then? 778 00:29:44,570 --> 00:29:46,729 Well, the value in 779 00:29:46,730 --> 00:29:49,309 court, I think, is that 780 00:29:49,310 --> 00:29:51,529 it's not really of value in court 781 00:29:51,530 --> 00:29:53,569 because it does not identify you. 782 00:29:53,570 --> 00:29:55,609 It's a value during the investigation. 783 00:29:55,610 --> 00:29:57,739 So on the very beginning of starting to 784 00:29:57,740 --> 00:29:59,839 look for who is 785 00:29:59,840 --> 00:30:01,490 responsible for the crime. 786 00:30:03,260 --> 00:30:04,999 Therefore, in the end, then you would 787 00:30:05,000 --> 00:30:07,279 still have the DNA trace of the 788 00:30:07,280 --> 00:30:09,019 DNA fingerprint and then you can still 789 00:30:09,020 --> 00:30:11,269 compare this to 790 00:30:11,270 --> 00:30:13,729 the person that you have found. 791 00:30:13,730 --> 00:30:15,829 And if this doesn't match, 792 00:30:15,830 --> 00:30:18,189 then this person could go free, 793 00:30:18,190 --> 00:30:19,279 hopefully. 794 00:30:19,280 --> 00:30:20,280 And 795 00:30:21,410 --> 00:30:24,259 comparing the probabilities 796 00:30:24,260 --> 00:30:26,959 here or multiplying them, 797 00:30:26,960 --> 00:30:29,059 then you just 798 00:30:29,060 --> 00:30:31,699 you just narrow down on something. 799 00:30:31,700 --> 00:30:33,949 He will never be able to knauer down 800 00:30:33,950 --> 00:30:36,169 on just two 801 00:30:36,170 --> 00:30:39,109 persons and then they are 802 00:30:39,110 --> 00:30:41,059 you just have to pick one or test them 803 00:30:41,060 --> 00:30:43,639 both for their DNA profiles. 804 00:30:43,640 --> 00:30:45,889 So I would be careful here 805 00:30:45,890 --> 00:30:48,140 to just multiply the 806 00:30:49,550 --> 00:30:50,550 probabilities. 807 00:30:52,090 --> 00:30:53,749 You ask questions. 808 00:30:53,750 --> 00:30:55,639 And if not, then a big round of applause 809 00:30:55,640 --> 00:30:56,640 for a greater.