Here Is a Human Being

by Misha Angrist

  By late summer the markets had rebounded enough to allow the company to close its fourth round: $45 million, which brought its total fund-raising to $91 million.27 Early-access customers had begun queuing up for $20,000 genomes (the $5,000 price tag was still some months away). “For us, sequencing a human genome these days is almost trivial,” said Drmanac.28

  Well then, I said, will Complete Genomics sequence the PGP-10? George had already told me that the company would,29 but I wanted to hear it from the horse’s mouth. “We will do it in our spare time,” Drmanac promised.30 Twenty-five years in, his love affair with DNA sequencing was still in full bloom. “There is something magical about a complete genome,” Rade said right before he rushed off to meet with the VCs. “Being able to say, ‘That’s all of it!'”31

  Seven miles up the road, Pacific Biosciences was trying to make its own magic. Founded in 2004, PacBio had gone about its R&D business quietly. But in early 2008, the company’s CEO, Hugh Martin, told the New York Times, “When we’re ready, we’re just going to win the X Prize.”32 That same week, at the Marco Island genome technology meeting, PacBio stunned a packed house with a presentation of its Single Molecule Real Time (SMRT) sequencing system. Observers tossed around adjectives like creative, exciting, thrilling, and dramatic.33 To close the meeting, the company sponsored fireworks on the Gulf Coast beach. Steve Turner, PacBio cofounder and chief technology officer, predicted that within a few years the company would be able to deliver complete and accurate human genomes in less than an hour.34

  Turner, however, is hardly the overbearing type. “I’m a huge Steve Turner fan,” Chad Nusbaum told me. “Whether the technology succeeds or not, I’m very impressed by him. He’s smart, creative, and resourceful, but also quite accessible as a human being. He’s not full of himself.”35 When we spoke on the phone I gave Turner every opportunity to take shots at his competitors, but he politely declined. Indeed, he was somewhat reluctant to talk to me at all, if only to honor his more superstitious coworkers. “There is a very famous Silicon Valley flop,” he said. “The Seg-way. Those guys were going to be Google-esque in their success. They hired a biographer who was going to be their official documenter of this historically important invention. But then of course the company failed miserably. Some people here worry we will jinx ourselves by talking.”36

  As a grad student at Cornell, Turner’s pal Jonas Korlach had an idea: What if one could actually watch a single molecule of DNA polymerase synthesizing DNA in real time? Turner was already working on nanostructures with the goal of finding better ways to manipulate and visualize DNA. He and Korlach realized that if they succeeded, then they would also be able to “see” which base was incorporated at any given time; that is, they would have the most powerful DNA sequencer ever. But how to bring it to fruition? Their technological breakthrough was something called a zero-mode waveguide, essentially a tiny, glass-bottomed well with metal sides—the whole thing is only a few dozen nanometers wide (about the size of a single virus). When a laser is shone at the ZMW from below, enough light gets in to visualize a single DNA polymerase molecule clutching a single nucleotide (a single “letter” of DNA) but with almost none of the surrounding noise.37

  SMRT sequencing, if it worked as expected, would be cheaper, longer, and much faster. It would be cheaper because by focusing on single DNA molecules, it would require very little in the way of chemical reagents to make it go. The machine itself would cost no more than most of the current crop DNA sequencers: probably $500,000–$600,000 (it eventually came online at $695,000). SMRT sequencing reads would be longer because unlike most other methods, the process would not actively terminate enzyme activity in order to build a chain; instead one would simply “feed” the enzyme nucleotides and then let the big dog run. “Our view is that this enzyme [DNA polymerase] is really a sequencing instrument in and of itself, and what a horrible shame to throw it away after every base you sequence,” Turner told Chemical & Engineering News. “If we free it up to do what nature has programmed it over billions of years of evolution to do, we can get the extraordinary features that it has of extreme frugality and high speed.”38

  By early 2009, PacBio had gotten its average read-length up to 946 bases and shown the ability to produce reads of greater than three thousand bases.39 One of the initial problems with most of the next-generation sequencing technologies—including the Polonator—had been short read lengths: in order to reconstruct a big, complicated genome with the newfangled machines, one had to piece together millions of short fragments against a reference sequence and repeat the sequence many, many times to make sure it was accurate. As I’ve noted earlier, this was like doing a jigsaw puzzle with millions of tiny pieces, some of which were indistinguishable from each other.* PacBio aimed to increase the size of the pieces, reduce their number, and therefore reduce the difficulty with which they fit together. Eventually the company hoped to generate reads of tens of thousands of bases in length—two to three orders of magnitude more contiguous DNA per read than the current state of the art.40 Finally, SMRT sequencing would be faster because, left to its own devices, DNA polymerase works fast. In its coming-out paper in Science, PacBio showed an average sequencing rate of five bases per second;41 Turner hoped to multiply that by a factor ten. If he could. And if he could put a million or more wells on a plate (early versions used just a few thousand; the first one was slated to have eighty thousand), then SMRT sequencing could read 100 billion bases an hour: That would mean a complete human genome could be read fifteen times over in fifteen minutes.42

  At the Cold Spring Harbor Personal Genomes Meeting, with commercial launch presumably less than a year away, Turner was wary of me. When I asked him to describe what PacBio’s machine looked like, he said I should look at an article in the next issue of Forbes,43 but conceded that it would only show the instrument with the covers off, not fully assembled.

  “If I am considering buying a Tesla,” I said, not mentioning that his machine would be more than four times the cost of the electric sports car, “of course I want to see the engine, but, you know, I also want to see the lines.”

  “I hear you,” he said. “Aesthetics are important. But we’re not at the right time for that relative to when you’ll be able to go to the website and click ‘add to cart.’ ”

  When I observed that PacBio had raised more than $260 million since 2004 and noted that that was a shitload of money by any standard, his boyish face remained impassive.44 He shrugged and stuck to the script.

  “We are in an all-out sprint before our machine goes commercial. We are following a schedule that has been tightly choreographed.”

  By the end of the year the company had at least a dozen prototypes and six early-access collaborators. It had begun to assemble a sales force, and it had floated the idea of an initial public offering in 2010.45 As I poked around, a few (jealous?) competitors harped on the low raw accuracy of PacBio’s sequence data; even some would-be customers said, in effect, “Come on already—enough with the fireworks, we want to see human genomes.” But the buzz persisted. Asked what sequencing technology they were most excited about, labs responding to an In Sequence survey mentioned PacBio most often.46 The company had had extraordinary success in raising both money and expectations.

  George, unsurprisingly, was on the scientific advisory boards of both PacBio and Complete Genomics. “They both have cultures I would love to be in,” he said. “They are almost academic places, but they have all the money academics don’t have. I’m almost ready to quit my day job. They’re both front-runners in DNA sequencing and I feel a lot more comfortable being affiliated with both than just one,” he said diplomatically. And then he gave me the smile. “Of course it could be there’s a third one that wins … It could be the Polonator!”47

  Francis Collins had undergone a metamorphosis.

  I first suspected something was up when I watched the webcast of him moderating the ELSI panel at the 2008 Biology of Genomes meeting at Cold Spring Harbor. Th
e subject was Direct-to-Consumer Marketing of Genomic Tests. The panelists were policy expert Kathy Hudson, National Coalition for Health Professional Education in Genetics director Joe McInerney, and, in their first joint public appearance, the Big Three: Kari Stefansson from deCODE, Dietrich Stephan from Navigenics, and Linda

  Avey from 23andMe. Faced with an audience of skeptical genome scientists, the company reps were both extending a hand (“we want to work together with you,” “we want to be regulated”) and passionate about their model. “I’m actually convinced,” said Stefansson in his smooth Icelandic accent, “that in the near future, genetic profiling like [our companies are] marketing is going to power the paradigm shift from interventional medicine to preventive medicine… . I think it is always laudable when people learn more about themselves.”48

  After the panelists had spoken, Collins served as interlocutor. His questions were incisive, sincere, and did not betray the anxieties I’d sensed from him in this same auditorium a year earlier. “Are these tests clinically useful?” “Should people be able to get the information whether or not there is a clinical intervention?” “Do we need a database and where should it live?”49 He seemed to me to be buoyant, open-minded, and fully engaged. I thought maybe this was due to the imminent passage of GINA. But that wasn’t the whole story. Three weeks later he announced he would step down as director of NHGRI.50 He was coy about his plans. The genome community was long on speculation: What was Francis up to?

  A few days after he made the announcement I sat on a panel with him and several others at the World Science Festival at New York University.51 The WSF was organized by physicist/PBS rock star Brian Greene and his wife, TV producer Tracy Day. It featured events all over town, a street fair, films, lectures, and robots shooting hoops.52 Saturday morning I took the train from my sister-in-law’s in Brooklyn to NYU and went to the breakfast buffet for speakers. As I was sitting there with my muffin and fruit cup, an elderly woman approached and asked if she could join me. It was Vera Rubin, perhaps the most decorated female astronomer in history, a former Richard Feynman student,53 and an all-around lovely person. This was pretty cool.

  I went out to the street fair. There I was accosted by self-styled revolutionary communists, who were pimping copies of their pro-evolution, anti-religion tract. I chatted with a scraggly guy who smelled faintly of alcohol. He tried to reassure me that communists didn’t think Islam was any worse than any other religion, despite my not having suggested otherwise. “They are all equally bad,” he insisted. He continued to press his book on me, but I took my leave—there was a terribly cute animatronic dinosaur I had to see.

  Down the street I bumped into friend and fellow panelist Jim Evans, a folksy and perspicacious medical geneticist at UNC and editor of Genetics in Medicine. He was curious about consumer genomics, so I had sent him my Navigenics report; he was singularly unimpressed (“Nothing new here. Did this really cost $2500?”).54 I noticed that he was wearing his DNA tie, the same tie I wore at my wedding.

  Later I entered the auditorium as Jim and the other participants in our session, “Your Biological Biography,” drifted in. Sociologist Nikolas Rose was there. I wanted to tell him what a big fan I was and that I had recommended him for this panel, but thought that might be poor form. Our moderator introduced himself. He was Sir Paul Nurse: London-born president of Rockefeller University, Nobel Prize–winning cell biologist, and extremely funny man. He reminded me of Dudley Moore—short and with the same shaggy mop top. He wore a dinner jacket over a black T-shirt emblazoned with the innards of the human torso. Backstage he told us his own ancestry story: A year earlier he had been detained by homeland security officials because his birth record documentation was somehow lacking in their eyes. He subsequently wrote to the hospital where he was born. His birth certificate arrived and in the space for “Mother” was his sister’s name while the space for “Father” was left blank. His presumptive parents were not his parents. He still didn’t know who his biological father was. He dryly observed that it “all came back to personal genomics, didn’t it?”55 “I’m not a bad geneticist,” he would say later, “but my own rather simple family kept a genetic secret from me for more than half a century.”56

  “Dr. Collins,” I said to Francis and extended a hand. He was nothing but cordial. His decision to walk away from NHGRI seemed to have lifted a huge weight from his shoulders. He asked me if I was writing a book. Had he forgotten our tense tête-à-tête the year before? As we waited in the wings to go onstage, I asked him what his plans were. He said he didn’t know—he might just enjoy being unemployed for a while. Or he might write his own book about personalized medicine.57

  He and I would meet again in a few months at the Personal Genomes meeting in Cold Spring Harbor. We interviewed each other over breakfast and this time he let me record him with no preconditions and nothing off the record. Unfettered by NIH (for the moment anyway), he was completely open.

  On his current thinking on personal genomics: “We have a long way to go before we can say that all of this is going to provide serious positive-outcome opportunities as opposed to just ‘curiosity satisfactions.’ But I guess I’m more on the side of, ‘Let’s go full speed ahead and empower the people interested in having this information sooner rather than later while still making sure we’re providing the kind of support they’re going to need.’ ”58

  Why was he so optimistic? “The most significant reason is that the science has finally gotten to the place where the predictions are not fanciful. Admittedly they are relatively modest risks [in most cases], but they’re not made-up. The other thing that’s given me an increased sense of confidence is GINA. So we have better science, we’re giving out better information, and we have better protections against the most egregious misuses. That really changes the landscape. I think that’s why I feel primarily positive versus two or three years ago when I felt primarily negative.”59

  I suggested that, given these developments, the current medical genetics paradigm was not up to the task of dealing with personal genomics and all that it entailed. To my surprise, he agreed. “We know that most common illnesses have heritability that’s in the neighborhood of fifty percent. If we had all [of those genetic factors in hand] then the idea that we could just go on with business as usual in medical genetics, which is largely built upon rare Mendelian conditions and chromosomal disorders, is not going to be sustainable. Frankly I’m very worried about my own specialty of medical genetics, a field I’m deeply attached to.”60

  Why has medical genetics not kept up? “We have not succeeded in attracting young, energetic, visionary physician scientists to join us. Here we are on the brink of a revolution where genetics is driving the entire medical profession, yet there are fewer applicants today to medical genetics fellowships than there were twenty years ago! Medical students look around and they don’t see a lot of geneticist role models because there are so few of us. We’ve never been able to catch the wave and build it up. I’m not sure the leaders [in medical genetics], most of whom were trained in an earlier era, are actually all that comfortable with the idea that they need to blow up their profession and start over.”61

  What would you do to fix it? “For starters we really have to renovate the training programs to make them attractive to the best and brightest of the next generation of physician-scientists. [Genetics] ought to be the kind of profession that lots of cerebral young physicians would like to join. Right now we’re doing what we’ve done for the last fifty years: [Trainees] see newborns with birth defects and go to the clinic and see people with [rare] Mendelian conditions. There’s not a sense that this is the new genetics—this is the old genetics. I think the medical genetics profession [has] more people who are ready to deal with dysmorphology [assessing people with congenital birth defects] than are ready to deal with a genomic sequence that needs to be sorted out. There’s still a ‘two cultures’ problem.”62

  I asked him about “the incidentalome.” What happens whe
n people learn potentially worrisome things about their genomes and start to pursue them at great cost and effort only to find out that most are meaningless? Collins said that that was a serious concern. “That’s the part I’m most wondering about when it comes to complete genome sequences. For these rare variants it will take decades before we have a sufficient ‘encyclopedia’ of them to know whether they matter. [In the meantime] we’ll be leaving a lot of ambiguity hanging in the air.”63

  We were running out of time before the morning sessions of the meeting would commence. I had caught Collins at a professional inflection point, but his embrace of personal genomics was only the beginning. In the coming months he would establish a foundation meant to address the culture war between science and faith in the United States.64 He would indeed write a book.65 In July 2009 he was nominated for the directorship of the NIH.66 A few weeks later he was confirmed unanimously by the Senate.67 Had his decision to resign from NHGRI been part of some master plan to become top boss at NIH?

  Shortly after his confirmation, he laid out his priorities. First on the list was using hi-tech approaches to discover the genetic bases of diseases.68 Whatever one’s view of Collins, his confirmation was unequivocal proof that genomics had reached the head of the class. And I expected it to stay there: Collins has long been interested in a large cohort study of genes, traits, and environments in the U.S. population akin to the UK Biobank.69

  There was another, smaller reason I was optimistic that Francis Collins—onetime would-be villain of this book—would do right not only by genomics writ large, but by our flagging health-care system and those who have been ill-served by it. A week after he was nominated to lead NIH I found myself in Washington, D.C., for Genetics Day on the Hill, a day of gentle lobbying of our elected representatives on issues pertaining to genetics or, as was the case in 2009, health care in general. We were given talking points—mostly of the mom-and-apple-pie variety—and we were broken up into teams of three to six people with whom we would visit our senators and congresspeople.