WEBVTT 00:00:00.000 --> 00:00:03.279 You know, usually when you think about a Black 00:00:03.279 --> 00:00:06.360 Friday door buster deal, you're picturing like 00:00:06.440 --> 00:00:10.140 a heavily discounted flat screen TV or maybe 00:00:10.140 --> 00:00:12.759 a new tablet. Right. Some kind of consumer electronics. 00:00:12.900 --> 00:00:16.059 Yeah, exactly. Something mass produced in a factory 00:00:16.059 --> 00:00:19.359 overseas that you just click add to cart on without 00:00:19.359 --> 00:00:21.480 really thinking twice. The kind of stuff that's 00:00:21.480 --> 00:00:23.500 designed to be cheap and disposable, honestly. 00:00:23.579 --> 00:00:26.079 Exactly. But you are definitely not picturing 00:00:26.079 --> 00:00:28.820 the fundamental biological blueprint of your 00:00:28.820 --> 00:00:31.320 own existence being tossed into an online shopping 00:00:31.320 --> 00:00:34.039 cart right next to like a pair of noise canceling 00:00:34.039 --> 00:00:36.929 headphones. Which sounds Completely absurd when 00:00:36.929 --> 00:00:39.369 you put it that way. It really does. But welcome 00:00:39.369 --> 00:00:42.049 to the deep dive. We've been pouring over a massive 00:00:42.049 --> 00:00:44.590 stack of source material today, specifically 00:00:44.590 --> 00:00:47.649 a comprehensive Wikipedia article and some historical 00:00:47.649 --> 00:00:50.789 data detailing the economics and, well, the wild 00:00:50.789 --> 00:00:53.329 commercial arms race of genome sequencing. It's 00:00:53.329 --> 00:00:55.969 a quiet story. It is. Our goal for you, the listener, 00:00:56.229 --> 00:00:58.409 is to figure out how reading human DNA went from 00:00:58.409 --> 00:01:01.289 a multi -billion dollar government mega project 00:01:01.289 --> 00:01:05.040 to literally a holiday discount deal. And more 00:01:05.040 --> 00:01:07.340 importantly, what that plummeting cost actually 00:01:07.340 --> 00:01:09.819 means for your health care going forward. Because 00:01:09.819 --> 00:01:12.560 it's moving fast. I mean, the speed of technology 00:01:12.560 --> 00:01:14.959 is just mind -blowing sometimes. It really is 00:01:14.959 --> 00:01:16.959 one of the most underappreciated technological 00:01:16.959 --> 00:01:20.349 leaps of our entire lifetime. but to really grasp 00:01:20.349 --> 00:01:23.209 the scale of what happened, we need to define 00:01:23.209 --> 00:01:25.810 the core concept that drove this whole race. 00:01:26.010 --> 00:01:29.590 Okay, lay it out for us. So, the phrase $1 ,000 00:01:29.590 --> 00:01:32.010 genome gets thrown around a lot in these sources, 00:01:32.510 --> 00:01:35.469 but it isn't just a random price tag. It represents 00:01:35.469 --> 00:01:38.549 a very specific, almost mythical threshold in 00:01:38.549 --> 00:01:40.510 medical history. Like breaking the four -minute 00:01:40.510 --> 00:01:43.030 mile. Exactly like that. The goal was to reach 00:01:43.030 --> 00:01:45.730 a point where fully sequencing an individual's 00:01:45.730 --> 00:01:48.019 whole genome, meaning reading all three billion 00:01:48.019 --> 00:01:51.719 base pairs of your DNA costs roughly $1 ,000. 00:01:52.239 --> 00:01:54.599 And why 1 ,000 specifically? Because at that 00:01:54.599 --> 00:01:57.359 price point, you cross a critical line. You take 00:01:57.359 --> 00:01:59.120 a scientific marvel that was only accessible 00:01:59.120 --> 00:02:01.459 to elite researchers, and you turn it into a 00:02:01.459 --> 00:02:03.939 routine medical test. Right. It is the absolute 00:02:03.939 --> 00:02:06.040 prerequisite for what we call predictive and 00:02:06.040 --> 00:02:08.620 personalized medicine. OK, let's unpack this. 00:02:09.099 --> 00:02:13.639 Because to understand why $1 ,000 is such a paradigm 00:02:13.639 --> 00:02:16.379 shifting number, We have to look at the original 00:02:16.379 --> 00:02:18.479 baseline, the starting line for this whole race. 00:02:18.580 --> 00:02:20.539 Right, we have to go back a bit. Yeah, back to 00:02:20.539 --> 00:02:24.159 the year 2001. The Human Genome Project publishes 00:02:24.159 --> 00:02:27.300 its very first draft of the Human Genetic Code 00:02:27.300 --> 00:02:31.360 and the cost of that achievement, a staggering 00:02:31.360 --> 00:02:36.110 $2 .7 billion. Yep. 2 .7 billion. And it took 00:02:36.110 --> 00:02:38.530 an entire decade to pull off. So going from a 00:02:38.530 --> 00:02:41.990 $2 .7 billion project to a $1 ,000 product in 00:02:41.990 --> 00:02:43.789 just a couple of decades, it's like spending 00:02:43.789 --> 00:02:45.909 billions to build the first Apollo rocket and 00:02:45.909 --> 00:02:47.650 then almost immediately demanding that a ticket 00:02:47.650 --> 00:02:49.949 to the moon should cost less than a laptop. I 00:02:49.949 --> 00:02:51.669 love that analogy. It's incredibly accurate, 00:02:51.810 --> 00:02:53.469 especially when you consider the sheer brute 00:02:53.469 --> 00:02:55.689 force required for that first draft. They really 00:02:55.689 --> 00:02:57.530 had to brute force it, didn't they? Oh, absolutely. 00:02:57.830 --> 00:03:00.770 The Human Genome Project relied on an older technology 00:03:00.770 --> 00:03:04.330 called Sanger sequencing. Imagine taking a massive 00:03:04.330 --> 00:03:06.810 multi -volume encyclopedia and reading it letter 00:03:06.810 --> 00:03:09.349 by letter, word by word, from cover to cover. 00:03:09.729 --> 00:03:12.849 Ugh, that sounds agonizingly slow. It was. They 00:03:12.849 --> 00:03:15.830 had entire warehouses filled with these massive 00:03:15.830 --> 00:03:18.610 machines running around the clock for years just 00:03:18.610 --> 00:03:22.270 to get one incomplete draft. Wow. So the scientists 00:03:22.270 --> 00:03:24.189 involved, you know, they knew they had just accomplished 00:03:24.189 --> 00:03:27.849 a monumental milestone for humanity. But they 00:03:27.849 --> 00:03:31.580 also recognized a massive glaring problem. Which 00:03:31.580 --> 00:03:35.319 is that $2 .7 billion is basically a national 00:03:35.319 --> 00:03:38.520 budget. Precisely. It is not a viable healthcare 00:03:38.520 --> 00:03:40.819 strategy for everyday people. Right. I mean, 00:03:40.979 --> 00:03:43.800 if my doctor needs a $2 .7 billion grant just 00:03:43.800 --> 00:03:46.300 to figure out why my knee hurts or what medication 00:03:46.300 --> 00:03:48.280 I should take, I'm probably just going to take 00:03:48.280 --> 00:03:50.539 my chances with ibuprofen. Exactly the problem. 00:03:50.800 --> 00:03:52.819 The technology had to become millions of times 00:03:52.819 --> 00:03:56.490 cheaper. So the phrase $1 ,000 genome was actually 00:03:56.490 --> 00:03:58.409 coined right at the tail end of that project. 00:03:58.629 --> 00:04:01.830 When was this? December 2001 at a National Human 00:04:01.830 --> 00:04:04.629 Genome Research Institute or NHGRI scientific 00:04:04.629 --> 00:04:07.169 retreat in Virginia. So they just planted a flag. 00:04:07.349 --> 00:04:09.889 They planted a flag and by 2002 heavyweights 00:04:09.889 --> 00:04:12.669 in the field like Jake Venter were aggressively 00:04:12.669 --> 00:04:14.870 pushing this goal at sequencing conferences. 00:04:15.009 --> 00:04:16.730 And didn't Venter actually put his money where 00:04:16.730 --> 00:04:20.220 his mouth is? He did. In 2003, his foundation 00:04:20.220 --> 00:04:21.879 put up half a million dollars for breakthrough 00:04:21.879 --> 00:04:24.740 to reach that price. And by 2004, the government 00:04:24.740 --> 00:04:27.920 formally stepped in. The NHGRI started handing 00:04:27.920 --> 00:04:30.560 out specific grants designed entirely to force 00:04:30.560 --> 00:04:32.300 the development of breakthrough technologies 00:04:32.300 --> 00:04:35.720 to hit this exact thousand dollar mark. It almost 00:04:35.720 --> 00:04:38.040 sounds like they were trying to mandate a technological 00:04:38.040 --> 00:04:40.720 miracle. But what was the ultimate vision here? 00:04:40.740 --> 00:04:43.079 Was the goal really just to have a cheap lab 00:04:43.079 --> 00:04:45.759 test? If we connect this to the bigger picture, 00:04:45.920 --> 00:04:49.139 it was much bigger than just a lab test. In a 00:04:49.139 --> 00:04:52.540 2006 article in Scientific American, the geneticist 00:04:52.540 --> 00:04:55.040 George M. Church laid out the vision beautifully. 00:04:55.220 --> 00:04:57.540 What did he say? He explained that the $1000 00:04:57.540 --> 00:05:00.079 genome was essentially shorthand for a promise. 00:05:00.389 --> 00:05:02.850 The promise that DNA sequencing would become 00:05:02.850 --> 00:05:05.750 so wildly affordable that an individual would 00:05:05.750 --> 00:05:07.769 view it as a once -in -a -lifetime expenditure. 00:05:08.149 --> 00:05:10.310 Oh, wow. So you just do it once. Exactly. You 00:05:10.310 --> 00:05:12.930 pay for it once, your full personal genome gets 00:05:12.930 --> 00:05:15.290 read onto a digital disk, and then your doctors 00:05:15.290 --> 00:05:17.089 just reference that disk for the rest of your 00:05:17.089 --> 00:05:18.790 life. Because your core code doesn't change. 00:05:19.089 --> 00:05:22.170 Right. It shifts the entire paradigm of medicine 00:05:22.170 --> 00:05:25.129 from reacting to your symptoms after you get 00:05:25.129 --> 00:05:28.480 sick to predicting your risks. based on your 00:05:28.480 --> 00:05:31.060 fundamental biology before anything ever happens. 00:05:31.560 --> 00:05:33.839 A once -in -a -lifetime medical purchase. That 00:05:33.839 --> 00:05:37.060 makes so much sense. So the vision is set. But 00:05:37.060 --> 00:05:39.740 having the vision is one thing. The source material 00:05:39.740 --> 00:05:42.300 highlights that actually pulling this off revealed 00:05:42.300 --> 00:05:45.480 a massive hidden roadblock. A very expensive 00:05:45.480 --> 00:05:47.779 roadblock. Yeah, because getting that data onto 00:05:47.779 --> 00:05:50.139 a disk turned out to be only half the battle. 00:05:50.240 --> 00:05:52.779 A very expensive half initially. The transition 00:05:52.779 --> 00:05:54.939 from that multi -billion dollar government project 00:05:54.939 --> 00:05:57.839 to individual sequencing started to happen, but 00:05:57.839 --> 00:05:59.959 it was incredibly clunky. Like that ceremony 00:05:59.959 --> 00:06:04.259 in 2007. Yes. May 2007, a highly publicized ceremony 00:06:04.259 --> 00:06:07.060 at Baylor College of Medicine. James D. Watson, 00:06:07.259 --> 00:06:09.560 the co -discoverer of the structure of DNA, was 00:06:09.560 --> 00:06:11.560 handed a digital copy of his personal genome. 00:06:11.740 --> 00:06:13.720 Right, they physically gave it to him on a portable 00:06:13.720 --> 00:06:16.600 hard drive. Which is so 2007. It really is. It 00:06:16.600 --> 00:06:19.240 was heralded as the first personal genome produced 00:06:19.240 --> 00:06:21.709 using next -generation sequencing. But I saw 00:06:21.709 --> 00:06:24.310 the price tag for that specific portable hard 00:06:24.310 --> 00:06:27.490 drive full of Watson's DNA. It was an estimated 00:06:27.490 --> 00:06:30.769 $1 million. Which, to be fair, is a massive discount 00:06:30.769 --> 00:06:33.670 from $2 .7 billion. True. But it's still a million 00:06:33.670 --> 00:06:36.850 dollars. And while the media was throwing a parade 00:06:36.850 --> 00:06:39.569 for this million dollar hard drive, the actual 00:06:39.569 --> 00:06:42.810 experts in the field were raising giant red flags. 00:06:42.889 --> 00:06:45.019 Oh, they were panicking. People like Bruce Korff, 00:06:45.139 --> 00:06:47.319 the past president of the American College of 00:06:47.319 --> 00:06:50.319 Medical Genetics, and Elaine Martis from Washington 00:06:50.319 --> 00:06:53.139 University, they started pointing out a crucial 00:06:53.139 --> 00:06:55.800 flaw in everyone's obsession with the cost of 00:06:55.800 --> 00:06:58.579 the sequencing itself. Right. Korff coined the 00:06:58.579 --> 00:07:02.120 phrase, the $1 million interpretation. And Martis 00:07:02.120 --> 00:07:06.339 called it the $100 ,000 analysis. But wait, I 00:07:06.339 --> 00:07:08.139 want to push back on this for a second. Go for 00:07:08.139 --> 00:07:10.519 it. If I hand my doctor a hard drive with my 00:07:10.519 --> 00:07:12.720 complete genetic code, isn't that the whole point? 00:07:13.069 --> 00:07:15.810 Why is the interpretation suddenly costing another 00:07:15.810 --> 00:07:18.610 million dollars? It feels like we're obsessing 00:07:18.610 --> 00:07:20.550 over the cost of printing the book, but what 00:07:20.550 --> 00:07:22.629 about the cost of actually reading the language? 00:07:22.970 --> 00:07:24.889 What's fascinating here is that you have hit 00:07:24.889 --> 00:07:28.310 on the exact bottleneck that terrified the medical 00:07:28.310 --> 00:07:31.850 community. Really? Yes. There is a massive distinction 00:07:31.850 --> 00:07:35.329 between generating the raw data, you know, just 00:07:35.329 --> 00:07:38.250 spitting out the billions of A's, C's, D's, and 00:07:38.250 --> 00:07:42.589 G's, and the immense computational labor required 00:07:42.589 --> 00:07:45.389 to actually analyze that data. So it's like paying 00:07:45.389 --> 00:07:48.250 someone a million dollars to perfectly transcribe 00:07:48.250 --> 00:07:51.490 a book written in an ancient dead language. That 00:07:51.490 --> 00:07:54.110 is brilliant, yes. Like you have the text perfectly 00:07:54.110 --> 00:07:57.290 copied down on paper, but it is completely useless 00:07:57.290 --> 00:08:00.410 until you hire an army of linguists at a hundred 00:08:00.410 --> 00:08:02.350 thousand dollars an hour to actually figure out 00:08:02.350 --> 00:08:04.689 what the story means. It perfectly captures the 00:08:04.689 --> 00:08:06.750 problem. You have three billion letters in your 00:08:06.750 --> 00:08:10.129 genome, but you also have millions of natural 00:08:10.129 --> 00:08:12.550 variations compared to a standard reference genome. 00:08:13.009 --> 00:08:15.209 OK. So let's say we find a spelling mistake on 00:08:15.209 --> 00:08:17.430 page four million of your code. What does that 00:08:17.430 --> 00:08:19.389 mistake actually mean? Right. Does it mean I'm 00:08:19.389 --> 00:08:21.589 going to get early onset Alzheimer's? Exactly. 00:08:21.889 --> 00:08:23.790 Or does it mean you're violently allergic to 00:08:23.790 --> 00:08:26.410 a certain type of common anesthesia? Or does 00:08:26.410 --> 00:08:28.449 it mean absolutely nothing and it's just a harmless 00:08:28.449 --> 00:08:31.430 quirk? And figuring that out isn't just a doctor 00:08:31.430 --> 00:08:34.440 squinting at a screen. Not at all. It requires 00:08:34.440 --> 00:08:37.620 supercomputers, proprietary algorithms, and teams 00:08:37.620 --> 00:08:41.019 of genetic counselors cross -referencing vast, 00:08:41.659 --> 00:08:44.200 constantly updating medical databases. Wow. So 00:08:44.200 --> 00:08:46.000 the physical sequencing was just the tip of the 00:08:46.000 --> 00:08:48.039 iceberg. The interpretation was where the real 00:08:48.039 --> 00:08:50.559 cost lay. So the implication there is terrifying 00:08:50.559 --> 00:08:52.639 for the industry. Because if the analysis and 00:08:52.639 --> 00:08:54.159 the supercomputers are going to cost you tens 00:08:54.159 --> 00:08:57.179 of thousands of dollars, the raw sequencing itself 00:08:57.179 --> 00:09:00.190 has to become practically free. Precisely. To 00:09:00.190 --> 00:09:02.549 make the whole package viable for a normal person, 00:09:03.110 --> 00:09:05.629 the raw data generation basically had to approach 00:09:05.629 --> 00:09:07.970 zero. It had to become a pure commodity. And 00:09:07.970 --> 00:09:09.970 nothing drives the price of a commodity down 00:09:09.970 --> 00:09:13.070 faster than a fiercely competitive commercial 00:09:13.070 --> 00:09:15.289 arms race. Which is exactly what kicked off in 00:09:15.289 --> 00:09:17.730 the biotech industry. Yeah, the late 2000s turned 00:09:17.730 --> 00:09:20.870 into a wild scramble to slash costs. At the end 00:09:20.870 --> 00:09:23.870 of 2007, a company called Gnome debuts a direct 00:09:23.870 --> 00:09:27.149 to consumer service. The price tag is $350 ,000. 00:09:27.649 --> 00:09:30.700 And wealthy individuals actually paid it. A Swiss 00:09:30.700 --> 00:09:33.340 biotech entrepreneur named Dan Stoichescu was 00:09:33.340 --> 00:09:36.149 one of their very first clients. That is wild. 00:09:36.309 --> 00:09:40.929 So in 2007, the floor is $350 ,000. But then 00:09:40.929 --> 00:09:43.090 the timeline just goes into free fall. It really 00:09:43.090 --> 00:09:46.509 does. In 2008, the company Illumina announces 00:09:46.509 --> 00:09:49.769 they can sequence an individual for $100 ,000 00:09:49.769 --> 00:09:52.409 in regent costs. Immediately, their competitor 00:09:52.409 --> 00:09:55.750 applied Biosystems counters with $60 ,000. Like 00:09:55.750 --> 00:09:59.190 a bidding war in reverse. Totally. By 2009, a 00:09:59.190 --> 00:10:01.730 Stanford professor named Stephen Quake sequences 00:10:01.730 --> 00:10:04.129 his own genome on a machine he co -founded for 00:10:04.129 --> 00:10:08.750 just I mean, how on earth is an industry slashing 00:10:08.750 --> 00:10:11.070 its prices in half every single year? Is this 00:10:11.070 --> 00:10:13.629 just Moore's law for biology? It is actually 00:10:13.629 --> 00:10:16.049 vastly outpacing Moore's Law. Really? Yeah, Moore's 00:10:16.049 --> 00:10:18.210 Law is the observation that computing power roughly 00:10:18.210 --> 00:10:21.649 doubles every two years. The drop in genome sequencing 00:10:21.649 --> 00:10:24.070 costs was falling much, much faster than that. 00:10:24.070 --> 00:10:26.870 It comes down to a fundamental shift in the technology. 00:10:27.250 --> 00:10:29.950 We move from that slow, page -by -page Sanger 00:10:29.950 --> 00:10:32.289 sequencing to what is called next -generation 00:10:32.289 --> 00:10:34.370 sequencing, or massively parallel sequencing. 00:10:34.490 --> 00:10:36.730 OK, how does massively parallel sequencing work? 00:10:36.950 --> 00:10:39.429 Because it clearly broke the cost curve entirely. 00:10:39.639 --> 00:10:42.259 Think back to our encyclopedia analogy. Instead 00:10:42.259 --> 00:10:44.720 of reading the 23 volumes of the human genome 00:10:44.720 --> 00:10:48.500 page by page, next generation sequencing takes 00:10:48.500 --> 00:10:52.059 30 complete copies of that encyclopedia set and 00:10:52.059 --> 00:10:54.320 runs them all through an industrial paper shredder. 00:10:54.440 --> 00:10:56.440 OK, wait. So now I just have a mountain of tiny 00:10:56.440 --> 00:10:59.679 paper strips. Right. Millions of tiny strips, 00:10:59.860 --> 00:11:01.919 each with just a few letters on them. That sounds 00:11:01.919 --> 00:11:04.919 chaotic. It is. But here's the trick. You dump 00:11:04.919 --> 00:11:07.940 all those strips onto a massive table. and chemical 00:11:07.940 --> 00:11:10.860 sensors read all of them simultaneously in parallel. 00:11:11.700 --> 00:11:13.899 Then you use a supercomputer to look for overlapping 00:11:13.899 --> 00:11:16.639 sentences on those strips to glue the whole puzzle 00:11:16.639 --> 00:11:19.000 back together. Because you are reading millions 00:11:19.000 --> 00:11:22.039 of strips at the exact same time, it is exponentially 00:11:22.039 --> 00:11:24.539 faster and cheaper. But wait, if you are shredding 00:11:24.539 --> 00:11:26.960 the encyclopedia, why do you need 30 copies of 00:11:26.960 --> 00:11:29.159 it to begin with? Why not just shred one copy? 00:11:29.259 --> 00:11:31.399 Because the shredding and reading process is 00:11:31.399 --> 00:11:34.460 incredibly messy. Sometimes a strip gets damaged, 00:11:34.919 --> 00:11:37.379 or the chemical sensor misreads a letter. If 00:11:37.379 --> 00:11:40.600 you only shredded one copy of the book, a mistake 00:11:40.600 --> 00:11:43.779 could ruin a crucial sentence. By shredding 30 00:11:43.779 --> 00:11:45.840 copies, you ensure that you have enough overlapping 00:11:45.840 --> 00:11:48.659 pieces to confidently catch errors and reconstruct 00:11:48.659 --> 00:11:50.460 the page perfectly. Okay, that makes perfect 00:11:50.460 --> 00:11:53.539 sense. That is a concept called 30x coverage. 00:11:54.360 --> 00:11:56.600 You are reading the equivalent of the whole genome 00:11:56.600 --> 00:11:59.519 30 times over to guarantee high accuracy. You 00:11:59.519 --> 00:12:01.779 need the redundancy because the cheap method 00:12:01.779 --> 00:12:04.960 is basically chaotic. Exactly. So this massively 00:12:04.960 --> 00:12:07.600 parallel shredded encyclopedia method is what 00:12:07.600 --> 00:12:11.240 drives the prices into the floor. By 2010, Illumina 00:12:11.240 --> 00:12:14.220 is offering a consumer service for $50 ,000. 00:12:14.620 --> 00:12:17.799 And their former CEO, John West, used it to sequence 00:12:17.799 --> 00:12:20.899 his entire family of four. Imagine dropping 200 00:12:20.899 --> 00:12:23.480 grand on a family lab test. But the trajectory 00:12:23.480 --> 00:12:26.679 keeps plunging until we hit January 2014. This 00:12:26.679 --> 00:12:28.879 is a massive moment in the source material. The 00:12:28.879 --> 00:12:31.340 huge milestone. Illumina launches a new machine 00:12:31.340 --> 00:12:34.539 called the HiSeq X10 sequencer, and they claim 00:12:34.539 --> 00:12:36.720 they have finally found the holy grail. They 00:12:36.720 --> 00:12:40.039 announced the first ever legitimate $1 ,000 genome 00:12:40.039 --> 00:12:43.559 at 30x coverage. The reaction to this announcement 00:12:43.559 --> 00:12:46.159 in the scientific community was absolutely electric. 00:12:46.980 --> 00:12:49.519 Michael Schatz, a highly respected researcher 00:12:49.519 --> 00:12:52.899 at Cold Spring Harbor Laboratory, famously compared 00:12:52.899 --> 00:12:55.360 this achievement to the development of the telescope 00:12:55.360 --> 00:12:58.759 or the microprocessor. Wow, those are huge comparisons. 00:12:58.940 --> 00:13:01.220 They are. It was heralded as a genuine turning 00:13:01.220 --> 00:13:04.159 point in human history. The telescope allowed 00:13:04.159 --> 00:13:07.659 us to see the infinite cosmos. The microprocessor 00:13:07.659 --> 00:13:10.159 allowed us to build the digital world. And the 00:13:10.159 --> 00:13:13.480 Hisak X10 allowed us to finally read our own 00:13:13.480 --> 00:13:15.960 microscopic blueprint affordably. Okay, here's 00:13:15.960 --> 00:13:19.049 where it gets really interesting, though. I am 00:13:19.049 --> 00:13:21.149 looking at what the critics said in the source 00:13:21.149 --> 00:13:23.149 material about this 2014 announcement, and I 00:13:23.149 --> 00:13:24.750 have to be the skeptic for a second. Please do. 00:13:24.929 --> 00:13:26.769 Because the fine print here is doing a lot of 00:13:26.769 --> 00:13:29.429 heavy lifting. Critics pointed out that to achieve 00:13:29.429 --> 00:13:32.710 this magical $1 ,000 number, a laboratory had 00:13:32.710 --> 00:13:35.750 to buy a minimum of 10 of these machines, which 00:13:35.750 --> 00:13:39.590 required a $10 million upfront investment. And 00:13:39.590 --> 00:13:42.509 that $1 ,000 calculation, it only factored in 00:13:42.509 --> 00:13:44.529 the chemical regions used for the test. Just 00:13:44.529 --> 00:13:47.409 the consumables, yeah. It completely ignored 00:13:47.409 --> 00:13:50.149 the electricity overhead to run these massive 00:13:50.149 --> 00:13:52.549 machines, the servers to store the data, the 00:13:52.549 --> 00:13:55.889 technicians to run them. Are we just looking 00:13:55.889 --> 00:13:58.149 at incredibly clever accounting here? It's a 00:13:58.149 --> 00:14:00.909 very fair critique. It sounds like when a company 00:14:00.909 --> 00:14:03.309 sells you a super cheap printer but gouges you 00:14:03.309 --> 00:14:05.889 on the ink, except in this case the printer costs 00:14:05.889 --> 00:14:07.789 10 million dollars and you have to buy 10 of 00:14:07.789 --> 00:14:10.450 them to get the deal. This raises an important 00:14:10.450 --> 00:14:12.710 question and you are highlighting a fundamental 00:14:12.710 --> 00:14:15.740 tension in this industry. There is a massive 00:14:15.740 --> 00:14:19.740 difference between raw regent costs, the chemicals 00:14:19.740 --> 00:14:22.840 used up during one specific run, and the total 00:14:22.840 --> 00:14:25.039 operational costs. Right, the actual cost of 00:14:25.039 --> 00:14:27.220 doing business. Exactly. You could not just walk 00:14:27.220 --> 00:14:29.320 into a clinic in 2014 with a thousand dollar 00:14:29.320 --> 00:14:31.860 bill and walk out with your genome. The clinic 00:14:31.860 --> 00:14:33.840 had to amortize that ten million dollar equipment 00:14:33.840 --> 00:14:36.899 cost, the massive power bills, and the salaries 00:14:36.899 --> 00:14:39.899 of the bioinformatics team across thousands and 00:14:39.899 --> 00:14:42.299 thousands of patients just to break even. So 00:14:42.299 --> 00:14:45.440 the $1 ,000 genome in 2014 was basically a marketing 00:14:45.440 --> 00:14:48.279 technicality. To a degree, yes. But even the 00:14:48.279 --> 00:14:50.960 harshest critics had to concede a larger truth, 00:14:51.159 --> 00:14:53.320 which was whether the true fully loaded cost 00:14:53.320 --> 00:14:58.659 in 2014 was $1 ,000 or $1 ,500. The sheer trajectory 00:14:58.659 --> 00:15:02.259 of the cost drop was undeniably historic. They 00:15:02.259 --> 00:15:04.580 had shaved literally millions of dollars off 00:15:04.580 --> 00:15:07.379 the price tag in less than a decade. The dam 00:15:07.379 --> 00:15:09.879 had broken. Wow. And the technology was advancing 00:15:09.879 --> 00:15:12.679 so rapidly that it actually started breaking 00:15:12.679 --> 00:15:14.799 the very incentive structures that were designed 00:15:14.799 --> 00:15:16.659 to encourage it in the first place. Wait. What 00:15:16.659 --> 00:15:18.580 do you mean by breaking the incentive structures? 00:15:18.720 --> 00:15:21.379 Have you ever heard of the Archon Genomics X 00:15:21.379 --> 00:15:25.080 Prize? I have. The X Prize Foundation is famous 00:15:25.080 --> 00:15:27.799 for putting up massive cash bounties for private 00:15:27.799 --> 00:15:29.340 spaceflight and things like that. Right. Well, 00:15:29.340 --> 00:15:31.460 they've put up a $10 million grand prize for 00:15:31.460 --> 00:15:34.279 genomics. The challenge was set for January 2013. 00:15:34.299 --> 00:15:37.000 OK. The prize would go to the first private team 00:15:37.000 --> 00:15:39.559 that could sequence 100 human genomes within 00:15:39.559 --> 00:15:43.159 30 days at a high accuracy and at an audited 00:15:43.159 --> 00:15:46.299 total cost of exactly $1 ,000 per genome. So 00:15:46.299 --> 00:15:47.799 it was supposed to be the ultimate catalyst, 00:15:48.039 --> 00:15:50.220 like a massive carrot dangled in front of the 00:15:50.220 --> 00:15:52.179 industry to force them to hit the target. That 00:15:52.179 --> 00:15:54.559 was the intent, but the reality of what happened 00:15:54.559 --> 00:15:56.340 is one of the most fascinating anecdotes from 00:15:56.340 --> 00:15:59.919 this entire era. What happened? In August 2013, 00:16:00.250 --> 00:16:02.549 Months before the competition was even supposed 00:16:02.549 --> 00:16:05.429 to officially conclude, the founders of the X 00:16:05.429 --> 00:16:09.610 Prize made a stunning announcement. They abruptly 00:16:09.610 --> 00:16:12.169 canceled the entire $10 million competition. 00:16:12.429 --> 00:16:14.830 They just pulled the plug on a $10 million prize. 00:16:15.029 --> 00:16:18.289 Why? Did no one enter? No. It was the exact opposite. 00:16:18.669 --> 00:16:20.990 They canceled it because, in their own words, 00:16:21.309 --> 00:16:24.129 the prize had been outpaced by innovation. You're 00:16:24.129 --> 00:16:27.110 kidding? I'm not. The founders realized that 00:16:27.110 --> 00:16:29.190 the private sector was moving so aggressively 00:16:29.190 --> 00:16:31.370 and the commercial arms race we just talked about 00:16:31.370 --> 00:16:34.450 was dropping costs so fast on its own that a 00:16:34.450 --> 00:16:36.649 10 million dollar prize was suddenly irrelevant. 00:16:36.840 --> 00:16:39.419 That is insane. He was no longer incentivizing 00:16:39.419 --> 00:16:41.620 the industry because the companies stood to make 00:16:41.620 --> 00:16:44.820 billions by capturing the market. The technological 00:16:44.820 --> 00:16:47.519 reality of the free market had literally outrun 00:16:47.519 --> 00:16:50.179 a 10 million dollar philanthropic moonshot. That 00:16:50.179 --> 00:16:52.580 is staggering. The market was moving so fast 00:16:52.580 --> 00:16:55.220 that a 10 million dollar bounty basically became 00:16:55.220 --> 00:16:57.860 pocket change to these companies. Exactly. And 00:16:57.860 --> 00:17:00.809 once that dam broke. The race to the bottom really 00:17:00.809 --> 00:17:03.950 accelerated. We transitioned away from a futuristic, 00:17:03.950 --> 00:17:06.970 elite medical procedure into the realm of aggressive 00:17:06.970 --> 00:17:09.049 retail marketing. This is where it gets surreal. 00:17:09.250 --> 00:17:12.509 Yeah. By 2017 and 2018, companies like Dante 00:17:12.509 --> 00:17:14.809 Labs in Europe and the Beijing Genomics Institute 00:17:14.809 --> 00:17:17.430 are pushing prices well under a thousand dollars. 00:17:18.009 --> 00:17:20.450 BGI was offering full sequencing for six hundred 00:17:20.450 --> 00:17:23.210 dollars in 2017. And then we hit the ultimate 00:17:23.210 --> 00:17:26.089 consumer shift, the moment where DNA officially 00:17:26.089 --> 00:17:29.609 becomes a retail commodity. July 28. on Amazon 00:17:29.609 --> 00:17:32.950 Prime Day. Yes. Dante Labs offered full, whole 00:17:32.950 --> 00:17:36.670 genome sequencing for $349. Amazon Prime Day. 00:17:36.750 --> 00:17:38.390 You're buying discounted Tupperware, a smart 00:17:38.390 --> 00:17:41.349 speaker, and your entire genetic code. It's hilarious. 00:17:41.490 --> 00:17:43.210 But the price war didn't stop there. November 00:17:43.210 --> 00:17:46.150 2018, Black Friday, Dante Labs drops the price 00:17:46.150 --> 00:17:49.349 to under $200. Unbelievable. And Veritas Genetics, 00:17:49.490 --> 00:17:51.230 which, by the way, was co -founded by George 00:17:51.230 --> 00:17:53.410 Church, the same guy who laid out the vision 00:17:53.410 --> 00:17:56.069 in Scientific American back in 2006. They offered 00:17:56.069 --> 00:17:59.190 whole genome sequencing for $199. It was limited 00:17:59.190 --> 00:18:00.990 to a thousand. thousand customers as a promotion, 00:18:01.289 --> 00:18:05.089 but still, $199. And geneticists note that the 00:18:05.089 --> 00:18:08.049 industry is now fiercely competing to hit a new, 00:18:08.170 --> 00:18:12.390 even lower target, the $100 genome. So what does 00:18:12.390 --> 00:18:15.089 this all mean for you? the listener. Let's take 00:18:15.089 --> 00:18:17.230 a step back and look at the sheer scale of what 00:18:17.230 --> 00:18:19.289 we've just discussed because the whiplash is 00:18:19.289 --> 00:18:21.809 intense. It really is. We literally went from 00:18:21.809 --> 00:18:24.450 the equivalent of an elite government supercomputer, 00:18:24.710 --> 00:18:27.490 a two point seven billion dollar endeavor that 00:18:27.490 --> 00:18:30.670 took a decade and thousands of scientists to 00:18:30.670 --> 00:18:33.809 an iPad Black Friday door buster in less than 00:18:33.809 --> 00:18:36.750 two decades. It means we have definitively entered 00:18:36.750 --> 00:18:39.650 the era of the one thousand dollar genome and 00:18:39.650 --> 00:18:42.289 honestly passed right through it. Yeah. The promise 00:18:42.289 --> 00:18:44.529 of the early two has been completely fulfilled. 00:18:44.829 --> 00:18:47.009 The physical act of reading human biology is 00:18:47.009 --> 00:18:49.630 no longer a financial barrier. The democratization 00:18:49.630 --> 00:18:51.849 of genetic data is here. Which is incredible 00:18:51.849 --> 00:18:54.289 for medicine. It is. It means your doctor can 00:18:54.289 --> 00:18:56.230 theoretically look at your code and know exactly 00:18:56.230 --> 00:18:58.190 which blood pressure medication will work for 00:18:58.190 --> 00:19:00.589 your specific biology and which one will give 00:19:00.589 --> 00:19:02.670 you adverse side effects. No more trial and error. 00:19:03.089 --> 00:19:06.930 Exactly. It means we can identify rare hereditary 00:19:06.930 --> 00:19:10.480 diseases before symptoms even show. It changes 00:19:10.480 --> 00:19:13.019 the entire foundation of how we treat human health. 00:19:13.259 --> 00:19:15.440 Right, because if the sequence costs less than 00:19:15.440 --> 00:19:17.880 a pair of sneakers, the business model of medicine 00:19:17.880 --> 00:19:20.869 has to change. We've traced this incredible journey 00:19:20.869 --> 00:19:25.029 today. We started with the massive $2 .7 billion 00:19:25.029 --> 00:19:28.509 human genome project and those warehouses of 00:19:28.509 --> 00:19:32.670 slow plotting machines. We watched James Watson 00:19:32.670 --> 00:19:35.650 get handed a $1 million hard drive that nobody 00:19:35.650 --> 00:19:38.349 could afford to actually interpret. We navigated 00:19:38.349 --> 00:19:40.710 the shredded encyclopedia breakthroughs of next 00:19:40.710 --> 00:19:43.150 -generation sequencing, the accounting debates 00:19:43.150 --> 00:19:45.950 of the $10 million aluminum machines in 2014. 00:19:46.130 --> 00:19:48.230 All of it. And we ended up clicking add to cart 00:19:48.230 --> 00:19:52.190 on a $199 Black Friday sale. This deep dive really 00:19:52.190 --> 00:19:54.349 highlights how astonishingly fast science can 00:19:54.349 --> 00:19:56.730 democratize our biology. The financial barrier 00:19:56.730 --> 00:19:58.990 to knowing exactly what you are made of is entirely 00:19:58.990 --> 00:20:01.549 gone. It is gone. But as we wrap up, I think 00:20:01.549 --> 00:20:03.809 there is a crucial lingering perspective we need 00:20:03.809 --> 00:20:05.410 to leave you with. OK, what is it? Throughout 00:20:05.410 --> 00:20:07.950 this entire journey, we have focused entirely 00:20:07.950 --> 00:20:11.250 on the race to read our DNA for pennies. But 00:20:11.250 --> 00:20:13.690 as the cost of generating that raw sequence approaches 00:20:13.690 --> 00:20:16.410 zero, we have to ask ourselves a much harder 00:20:16.410 --> 00:20:19.119 question about the future. Which is? In a world 00:20:19.119 --> 00:20:21.660 where your genetic data is practically free to 00:20:21.660 --> 00:20:25.000 extract, who will actually hold the power over 00:20:25.000 --> 00:20:28.400 your biological destiny? Will it be you? Or will 00:20:28.400 --> 00:20:30.019 it be the companies giving away the sequencing 00:20:30.019 --> 00:20:32.880 at rock bottom prices, simply so they can control, 00:20:33.019 --> 00:20:35.440 database, and monetize the immensely valuable 00:20:35.440 --> 00:20:38.220 and still very expensive algorithms that interpret 00:20:38.220 --> 00:20:38.759 who you are?