Why Life Science Needs Its Own Silicon Valley

Sometime soon, in some location on Planet Earth, an assortment of companies, research institutions, entrepreneurs, and scientists will cluster together in an industrial ecosystem. Their goal: to exploit the rapid discoveries about the human genome—the DNA template that drives the development of a person’s biological functions. This new Silicon Valley will give rise to radical new ways of diagnosing and treating disease. With skillful management and luck, it will dominate the field of genomics for many years, bringing jobs and prestige to the area that hosts it. But where will this ecosystem arise? The United States, the UK, and Canada are all well positioned to stake a claim. At this point it’s anyone’s guess as to which will prevail—or whether another player will come to the fore. We do know that the window of opportunity won’t remain open long and that as soon as one country or region makes a move, others will be at a serious disadvantage. The U.S. has a good start: It is creating a vast database of genetic information on military veterans. But more work is needed to turn that database into an effective tool for diagnosis and treatment and an enduring wellspring of innovation. Early Advances It took 13 years and almost $4 billion in government and private funding for researchers in the Human Genome Project to make the first complete map of a person’s genome. Today, just nine years after that breakthrough, a genome can be mapped in a matter of weeks for about $2,000—and a genome map for half that is within sight. For $1,000 a company in Iceland will chart your genetic propensity for 47 different diseases and traits, ranging from diabetes to male-pattern baldness. Beijing Genomics Institute is on track to produce 10,000 entire human genome sequences a year. Soon doctors will be able to incorporate genetic sequencing into routine care. Sequencing technology has already demonstrated its value. In the first diagnosis by DNA, in 2009, a team led by the Yale geneticist Richard Lifton learned that a genetic mutation was responsible for a five-year-old’s severe dehydration; the mutation caused a particular type of diarrhea, which was then treated through diet. And sequencing led to the discovery that seemingly identical breast cancers have different genetic profiles, paving the way for better-targeted treatments. There’s also the economic impact: New preventive measures will save patients, insurers, and employers money, and studies project that genomic medicine will generate $350 billion worth of economic activity and millions of jobs. According to the nonprofit research organization Battelle, the U.S. genetics-testing industry has already created more than 116,000 jobs and produced $16.5 billion in economic output. Still, we’re a long way from being able to fully use the data encoded in our chromosomes. Although the Human Genome Project has determined the sequences of the 3 billion chemical base pairs in human DNA, its map shows a reference genome, not your genome. And so far sequencing can diagnose only single-mutation diseases, which account for just 5% of ailments. The rest are caused by instructions coded in multiple genes or by interactions between genetic and environmental factors. Making Clusters Happen Getting to the point at which genome studies are widely useful will require significant strides in the new field of bio IT. That’s where clusters—first described by the Harvard Business School professor Michael Porter—enter the picture. Clusters of producers, suppliers, and training centers often arise when business segments require high levels of specialization from multiple contributors. The 19th-century jute industry was clustered around Dundee, Scotland. Surat, India, is home to a diamond-cutting cluster. Dongguan, China, is the site of an electronics cluster. Your German car, your Russian vodka, your Brazilian coffee—all are products of industry clusters. Clusters of Clusters Biotech industry clusters, and groups of clusters, have arisen around the world. Some may soon evolve into clusters focused more on the business of gene science. Source: Minnesota Biomedical and Bioscience Network The genomics cluster will include multinational corporations, research institutions, scientists, students, investors, related industries, and start-ups that haven’t been imagined yet. And although the IT part of bio IT relies heavily on the internet, geography will be a crucial factor: The genomics cluster will have a physical location. Studies show that having a high concentration of people working on similar problems in the same location speeds progress. People run into one another in hallways, cafés, and train stations, and during these encounters they often exchange ideas. The collaboration and inspiration necessary for innovation are much easier on the ground than in the cloud. Governments are well aware of the job-creating and economy-boosting value of clusters, and they work hard to foster them. But no one really knows how to create a cluster. There’s no validated theory of a cluster’s life cycle. Consider the question of location alone: Although certain geographical sites make immediate sense—the U.S. auto capital, Detroit, is near iron ranges—clusters can arise in unexpected places for idiosyncratic reasons. A software cluster probably wouldn’t have sprung up around Seattle if Bill Gates hadn’t been born there. One factor does seem to be important for cluster creation, however: government support. The movie industry in Vancouver, which employs almost 50,000 people and contributes more than $1 billion annually to British Columbia’s economy, has benefited enormously from tax incentives and other forms of government cooperation. But financial incentives are only part of the story. Governments can also make a difference by providing such things as infrastructure, demand for products, and incentives and avenues for knowledge building. In the 1990s, in support of a biotech cluster in Cambridge, England, the British government funded an intellectual infrastructure of sorts—a networking organization known as the Eastern Region Biotechnology Initiative, which provided a means for industry participants to meet and interact, according to Theo Papaioannou, of the Open University’s Innogen Centre. The cluster ultimately grew to include more than 200 companies, 30 research institutes, and four research hospitals. A smaller and more widely dispersed biotech cluster around Dundee (the jute industry faded away long ago) was supported by the Scottish government. No one really knows how to create a cluster. But government support is important. For genomics, the most important element a government could help provide is a coherent, accessible source of data. As the cost of sequencing drops and DNA analysis becomes increasingly commonplace, genomic data will rapidly accumulate. If the information is in a usable form, researchers will be able to mine it to identify the root causes of complex genetic diseases. But individual scientists and research institutions have few incentives to build data banks that merge their results with those of other researchers; they’re typically focused on publishing articles or patenting new processes (or even new organisms). A national government will have to play a big role. Amassing a Database A truly useful genomics database would include more than DNA findings. It would contain information on disease outbreaks, patients’ family histories and environmental exposures, and, ideally, even information about diet and lifestyle, gleaned, for example, from grocery-store receipts. Such a database would be an invaluable innovation engine—a catalyst for new insights about diseases and treatments. J. Craig Venter on Where Genomics May Be Headed One of the first scientists to sequence the human genome, Venter founded Celera Genomics and the Venter Institute and cofounded Synthetic Genomics. He recently spoke with HBR’s Andrew O’Connell. Has technology made geographic clusters less important? Although the internet means that researchers aren’t limited by geography anymore, the genomics industry will probably form in specific locations, because of a critical factor: the employee base. One of the reasons the Venter Institute located in La Jolla, California, is the highly skilled workforce there. Does La Jolla have a shot at becoming the Silicon Valley of genomics? Depending on your definition, you could argue that a cluster has already formed there. It includes UC San Diego, the Salk Institute, the Scripps Research Institute, the Venter Institute, Synthetic Genomics, and 30 or 40 companies. All are within a few square miles, and all are using genomics. What other locations look promising? A cluster is growing in the Boston area. The UK’s genomics activity is centered in Cambridge, but it’s relatively small. You can’t count out India, because of its computational expertise. China has the Beijing Genomics Institute, and the government hopes to generate related economic development—but at this point it’s a cluster of one. The U.S. Department of Veterans Affairs recently began recruiting veterans to contribute blood samples and health and lifestyle information to the Million Veteran Program, so named because the government hopes to enroll a million veterans—roughly 20% of those who use VA health services—over the next five to seven years. The purpose is to help researchers learn how genes affect health. According to Ronald M. Przygodzki, the acting director of the VA’s Biomedical Laboratory Research and Development office, a database that large will give researchers studying even rare diseases the statistical power to replicate and verify their results. This is no small feat. But if the U.S. wants to create a database that will foster the growth of a genomics cluster, it must gather information on populations other than veterans, and it must find a faster way of compiling, cataloguing, and verifying that information. We estimate that a million-person genomics database would require an investment of several billion dollars. A few nations and organizations around the world are beginning to take steps that could facilitate the creation of a significant genomics database. The government of Iceland has fostered the establishment of a national genomics data bank that draws on that country’s unusually extensive family records. The U.S. National Center for Biotechnology Information provides access to Human Genome Project data through its public website. The British Columbia Cancer Agency, the University of Manitoba, the UK’s Cambridge Research Institute, and numerous other organizations are collaborating on a $4.5 million breast cancer study that also aims to gather genomic data and make it available for use in developing drugs. A genomics cluster will need more than a database, of course. It will need government support for R&D and for the creation of research parks and other facilities. The United States has hosted a number of dynamic clusters in the past, but lately a few have slipped away: Witness the migration of computer chip manufacture to Taiwan and China, for example. If America has the vision to lay the groundwork for a genomics cluster, the result could be a compelling two-for-one deal—a new industry ecosystem that forever alters the course of medicine while bolstering the country’s economic health.

Why Life Science Needs Its Own Silicon Valley

Partner Center