There is an Errata (April 2003) associated with this Feature.
Decoding the literature on genetic variation
Roger Coronini, Marie-Angèle de Looze, Pierre Puget, Gérard Bley
& Shyama V. Ramani
Roger Coronini, Marie-Angèle de Looze, and Shyama V. Ramani are researchers at the Institut National de la Recherche Agronomique, Université Pierre Mendès France, Grenoble, France, and Pierre Puget and Gérard Bley are researchers at the Commissariat à l'Energie Atomique, Laboratoire d'Electronique et des Technologies de l'Information, Grenoble, France
A survey of the scientific and patent literature on single-nucleotide variants reveals the dominance of research centers in the United States and the prolific patenting of SNP technology by a select group of biotechnology companies.
Single-nucleotide polymorphisms (SNPs) are DNA sequence variations among individuals. Publicly funded laboratories and private businesses are attempting to associate specific SNPs (or sets of SNPs) with various medical conditions and to study the differences in SNP patterns among various human populations. Ultimately, it is hoped that knowledge of SNPs will improve medical treatment by enabling prediction of disease risk and response to therapies. To facilitate these efforts, the importance of providing publicly accessible SNP data without intellectual property restrictions prompted the formation of the SNP Consortium (http://snp.cshl.org/), a public-private initiative that, as Nature Biotechnology went to press, has now placed 1.8 million SNPs in the public arena.
In this article, we present an analysis of the patent and scientific literature on SNPs (up to the end of 2001) to identify the key academic researchers and centers of excellence in the area, assess the major commercial developers of SNP technology, and understand the nature and progress of work currently underway. To assess the creation of new knowledge, we conducted a detailed study of papers published in the scientific literature. We have also measured SNP technological innovation by assessing patent applications filed at the major patent offices (see "Methodology"). The survey covers the scientific and patent literature over the period 1987−2001.
An emerging area From 1987 to 2001, 1,828 papers were published and 365 patents were filed on the topic of SNPs. Of these totals, 82% of papers were published and 88% of patents were filed between 1998 and 2001, suggesting that the field is still relatively young (Fig. 1). Since 1998, the number of publications has increased fourfold, and the number of patents threefold—reflecting a sudden spurt of knowledge creation quite typical of an emerging area.
Figure 1. Scientific papers and patents relating to SNP research in the period 1987−2001.
As Figure 1 indicates, a sharp increase in the number of scientific papers and patent publications occurred almost simultaneously in 1998. This is noteworthy because in many high-technology sectors, spurts in scientific publication tend to precede increases in the number of patents filed by a few years. As SNPs are the latest in a long line of genetic markers (including minisatellites and microsatellites) used in efforts to map human disease and other complex traits, it is possible that familiarity with mapping technology facilitated their rapid incorporation into company R&D programs. Moreover, SNPs clearly are of interest to both university researchers and industrial researchers: for academics, they provide a means of mapping traits to genomes at higher and higher resolution; for industrial researchers, they are of use in designing SNP diagnostics (for instance, to determine disease risk or drug response) or in basic drug-discovery research to design chemical and biological entities that can effectively address all polymorphic variants of a drug target.
Patent applications are published only 18 months after deposition. Several more months pass before they are collated into databases—for example, the Derwent Biotechnology Abstracts database is updated every trimester, whereas the World Patents Index is updated every week. A similar cataloging delay occurs for scientific publications—the Biotechnology Citation Index is updated every trimester, whereas the Web of Science is updated daily. Of course, the process of peer-reviewed publication also introduces delays into the release of results. Our present study was conducted at the end of 2001, and thus the data relating to the years 2000 and 2001 are necessarily incomplete, which accounts for the lower numbers of patents between 2000 and 2001 (Fig. 1). This might also explain why some prominent biotechnology companies involved in SNP research—such as Luminex (Austin, TX) and Pyrosequencing AB (Uppsala, Sweden)—did not appear in our survey; their patents or publications may have appeared in the literature after the time period covered by our survey.
US papers predominate We have categorized authors of SNP papers in two ways: those who publish the greatest number of papers (Table 1) and those whose work is referred to most frequently in the scientific literature (Table 2). The "most cited" authors listed may be regarded as the researchers with the greatest impact on the work of their scientific peers, and can therefore be viewed as leaders with the most influence in the creation of knowledge.
Table 1. The most prolific authors of SNP-related papers
The convention that academic scientists disclose and share data in journals more readily than their corporate colleagues is supported by Table 1, which shows that the top ten "most prolific" authors of scientific papers on SNPs all come from public institutions. Unlike their industrial colleagues, academic scientists are judged and ranked according to their publication record and publish regularly to ensure future grant awards. Notably, the majority of authors in Table 1 come from research institutions in the United States and Japan, with two originating from European research centers. Although certain companies clearly opt against scientific publication, an examination of the 30 most prolific authors does, however, reveal some companies that do take this route; for example, the Canon Research Center (Kanagawa, Japan) and Perkin Elmer (now renamed Applied Biosystems; Foster City, CA) have published 10 and 8 papers on SNPs, respectively.
The pre-eminence of US researchers in the SNP field is confirmed by data presented in Table 2. Francis Collins heads the list with 219 citations: given his status as director of the US National Human Genome Research Institute (Rockville, MD), many of his papers presumably present policy and direction for SNP research efforts in the United States and are therefore highly cited. Data from Table 2 also indicate that the Whitehead Institute for Biomedical Research (Cambridge, MA) is a world-leading center for SNP research, providing a home for several top research groups. Other "most cited" authors come from laboratories elsewhere in the United States, in Japan, or in Sweden.
An examination of the 30 top-cited authors reveals that the US lead in the area of SNP research extends far beyond the top ten researchers. Of the 30 top-cited authors, 23 are located in the United States; of the other 7, 4 hail from Scandinavian countries (Sweden and Finland), and 1 each from Japan and the United Kingdom.
Ten researchers figure among the top 30 most prolific authors as well as the top 30 most cited authors. Again, most of these are from the United States; notably, Eric Lander and Deborah A. Nickerson are among the top ten in both categories. Only three European researchers are among the top 30 most cited authors: Ulf Landegren of the University of Uppsala (Uppsala, Sweden) with 75 citations, Anthony J. Brookes of the Karolinska Institute Centre (Stockholm, Sweden) with 50 citations, and Tomi Pastinen of the National Public Health Institute (Helsinki, Finland) with 73 citations. Although four Japanese researchers are among the 10 most prolific authors, they are not among the top 30 most cited.
Voracious biotech patenting In total, we identified 164 different organizations that applied for SNP patents in the time period 1987−2001. To identify organizations that have strategically focused their intellectual property on SNPs, Table 3 lists 30 patent applicants that have deposited at least three patents each (from now on, we refer to this group of patentees as "leaders in patenting"). Their 241 patent applications account for 66% of the total number of applications on SNPs.
Table 3. Leading patentees of SNP research: main characteristics
As one would expect, Table 3 reveals that a majority (18) of the 30 "leaders" involved in patenting SNPs are research-based biotechnology companies. Eleven of these 18 companies are startups of less than 10 years old. Clearly, as for SNP scientific papers, the patent literature on SNPs is dominated by US-based organizations. Many of these companies are developing technology platforms to transform SNP research into medical applications either for their own in-house drug discovery programs or to provide diagnostic and/or genotyping services to others. Commercializing kits for typing SNPs is also the focus for biotechnology companies focusing on reagents and equipment, such as Qiagen and Promega.
The rest of the patentees include a smattering of large pharmaceutical firms and some not-for-profit research centers. The use of SNPs in identifying traits of agronomic importance is emphasized by the intellectual property in this area held by Monsanto and Pioneer Hi-Bred. SNP patent applications in agriculture account for 2% of the total number of applications by the top 30 patentees.
Overall, research-based biotechnology companies have a total of 184 patent applications—or around 76% of all the patent applications of the "leaders." The supplier companies, pharmaceutical companies, and five not-for-profit research centers account for 17 (7%), 13 (7%), and 21 (7%) of the total, respectively.
It must be kept in mind that for many biotechnology companies, such as Curagen, Molecular Tool (acquired in 1998 by Orchid), Orchid Biocomputer (now Orchid Bioscience), and Affymetrix (and its spin-off Perlegen; Santa Clara, CA), the recent spurt in patent filings is the outcome of past investment in SNP research. These companies have been applying for SNP patents since the start of the 1990s and are now poised to be strategic players in this narrow market.
Twenty-five of the 30 patenting "leaders" are affiliated with US organizations (26 if you count Glaxo as part of Glaxo SmithKline). The remainder are linked with Swedish, UK, German, or French companies. Notably, the Japanese are absent among the top 30 patentees (even though Japan has ten patents in this domain, no single Japanese organization has as many as three patents to its credit).
Also noteworthy is the emphasis on patenting as compared to publishing among the different biotechnology companies listed in Table 3. With the exception of Aclara Biosciences, Affymetrix, Biogen, Illumina, Genset, and Curagen, biotechnology companies appear to prefer patent filing to publishing papers; indeed, two leading patentees, Molecular Tool (acquired by Orchid Biosciences), and Epigenomics AG), have no publications to their name. (As Isis Innovation is the technology-transfer arm of Oxford University, it is no surprise that it produces no publications.)
To a certain extent, the lower number of papers originating from companies may reflect the fact that a database like the Web of Science (which collates about 8,500 journals) is very selective, including only mainstream academic journals. But overall, the emphasis of private enterprises (particularly startup companies) toward patents (rather than papers) reflects the importance of product focus and trade secrets for maintaining competitive edge in the commercial market. Moreover, because the technology transfer arms of universities involved in filing for patents (such as the University of Alabama, Birmingham Research Foundation or the Wisconsin Alumni Research Foundation) are separate from the university institutions where research actually takes place, they also are unlikely to be linked with academics who author significant numbers of scientific papers (Table 3).
Ailments, apparatus, and alleles Table 4 lists the top ten public organizations and top ten companies in terms of scientific papers and patent applications. Clearly, those public organizations involved in human genome research with particular expertise in SNPs (e.g., the Stanford Human Genome Center, the University of Tokyo Human Genome Center, and the University of Washington Genome Center) have been most pro-lific. In contrast, very few biotechnology companies have published reports related to SNPs indexed by the Web of Science.
Table 4. Distribution of publications and patents in all areas of SNP research among top ten public organizations and companies*
In terms of patents, Curagen, Genaissance Pharmaceuticals, Orchid Biosciences (formerly Orchid Biocomputer, which also acquired Molecular Tool), and Incyte Genomics are clearly prolifically filing patents on SNPs. It is interesting to note that startup Genaissance Pharmaceuticals was founded in 1998 and originated out of Yale University, which is among the top ten public organizations publishing papers on SNPs.
For the sake of simplicity, we have further grouped our analysis of scientific publications and patents on SNPs into four main areas: associations of SNPs with human disease (Table 5), methods and techniques for scoring and discovering SNPs (Table 6), basic molecular biology of SNPs (Table 7), and allele frequencies of SNPs in populations (Table 8). In each area of SNPs, each Table presents public organizations who have published 10 or more publications on SNPs and patentees with 2 or more patent applications.
Table 5. Distribution of publications and patents focused on association of SNPs with human diseases and disorders
Taking into account all four areas, 1,481 papers (907 of them shown in Tables 5,6,7,8) appeared in the scientific literature in the period 1987−2001. The remaining 347 papers on SNPs focus on discussing the technology's "promise" in healthcare, agriculture, and basic biology, reflecting the fact that papers commenting on SNP research far outweigh, numerically, those detailing data of practical use.
SNP associations and disease More than a quarter of all biotechnology papers and patents concerning SNPs deal with their associations with complex disease (particularly those afflicting much of the developed world, such as arthritis, asthma, obesity, and cancer; Table 5). Researchers hope to detect the SNPs associated with predispositions to these diseases, which could lead them to the underlying genes. There have been several papers and patents in this area that focus on applying SNP information to diagnostics and nucleic acid therapies.
The majority of biotechnology papers in this area originate from public laboratories in the United States, but Humboldt University in Germany, the University of Utrecht in the Netherlands, Kagawa University (Kagawa, Japan), McGill University (Toronto, Canada), the University of Newcastle (Newcastle, UK), the University of Manchester (Manchester, UK), the University of Oxford (Oxford, UK), and Wakunaga Pharmaceuticals (Osaka) in Japan also are publishing in this area.
The large number of patent applica-tions in this area from Genaissance Pharmaceuticals and Curagen indicates that SNP associations with human disease form a strategic focus for these companies, with Orchid Biosciences also strong in the field. Several public organizations have been filing for patents in this area, including the Whitehead Institute for Biomedical Research, the University of Alabama Research Foundation (Birmingham, AL) and the University of California system. Interestingly, one company, PPGx (Research Triangle Park, NC), has five papers published in this area but no patents filed, according to our analysis. In addition to research-intensive biotechnology companies, well-established multinationals, such as AstraZeneca AB, have also been filing patents in regard to disease-related SNPs.
Methods and techniques By far the greatest proportion of SNP patent applications (55% of the total number of SNP patents) focus on technologies that enable high-throughput detection of genetic variants. Research in this area is published more as patents (55%) than as papers (18% of the total number of SNP papers), emphasizing the importance of this area to companies. Patents mostly describe protocols and biochemical methods (such as amplification methods, nucleic acids hybridization methods, or enzymatic methods), their implementation using innovative devices (such as microarrays or microfluidics systems), or ways to automate them (such as automated fluid handling and detection).
Leading biotechnology companies with a strategic focus on technology platforms for SNP genotyping or detection include Orchid Biosciences, Aclara Bioscience, Curagen, Epigenomics AG, Nanogen, Affymetrix, Incyte Genomics, and Illumina. Although these comprise by far the largest group of companies patenting SNPs tools, some large multinational pharmaceutical companies (such as AstraZeneca and GlaxoSmithKline) are also represented, although with many fewer applications. Some of the patentees are equipment and reagent suppliers and manufacturers, such as Perkin Elmer and Becton Dickinson; others are information technology companies that have moved into the life sciences, such as Hewlett Packard.
Public organizations active in this area include the Cornell Research Foundation, SNP consortium member the Whitehead Institute for Biomedical Research, the Wisconsin Alumni Research Foundation, the University of California system, and the University of Washington (Seattle). Notably, the Whitehead Institute and the University of Washington host several of the top ten cited authors listed in Table 2: Eric Lander, David Wang, Michele Cargill, and Leonid Kruglyak all are based at the Whitehead Institute, and Deborah Nickerson is on the faculty of the University of Washington.
The United States again is the most prominent country from which research on SNP tools is being published and patented. Nearly all (29 of 32) of the biotechnology companies working in this area—all except Epigenomics AG (Germany), Keygene (the Netherlands), and Asper Biotech (Tartu, Estonia)—are based in the United States, and only one European pharmaceutical company (AstraZeneca, Sweden) is patenting SNP tools.
Molecular biology of SNPs A small number of papers (21% of total) and patents (19% of total) focus on the effects of point mutations or sequence alterations such as shifts in bases, insertion, or substitution of amino acids, premature stop codons, change of splicing sites, and so on in a wide variety of genes (BRCA1 and p53 being prominent examples).
The main research centers publishing in this area include the National Cancer Center Research Institute of the University of Tokyo, the School of Medicine at Washington University (St. Louis), and the University of California system of campuses. Authors cited in Tables 1 and 2 include Yusuke Nakamura at the Human Genome Center of the University of Tokyo, Masato Orita of the National Cancer Center Research Institute, and Pui-Yan Kwok of the School of Medicine, Washington University.
Leading biotechnology research companies focusing on this area include Curagen, Incyte Pharmaceuticals, and Genaissance Pharmaceuticals. They also include older, more traditional biotechnology companies, such as Biogen, Genentech, and French genetics company Genset. Agbiotech pioneer Monsanto also has patents filed in this area.
Once again, US organizations are predominant, although Japan (for instance, the University of Tokyo) and Canada (McGill University) also have academic centers that are actively publishing papers in the area.
Allele frequencies of SNPs Papers describing differences in the frequency of certain sets of SNPs in populations either from specific geographical regions (for example, the United States) or representing specific races (for example, Caucasian) can be grouped together in a final area. From the data we analyzed, no patents have appeared in the literature corresponding to this area, perhaps because the practical implications for healthcare or agriculture are not clear or because assessing the genetic basis of "race" is rather controversial.
Public research organizations are much more heavily involved than private companies in studying the differences in SNPs among different populations, and the investment of the Japanese is particularly notable. For example, the group of Yusuke Nakamura, Katsushi Tokumaga, and Toshihiko Tanaka at the University of Tokyo published the most papers (12% of the total publications in this area) in our period of review (1987−2001). Public research institutions are more active than private companies in this field. As for all the other areas of SNP research, US researchers predominate in the scientific literature.
Conclusions SNPs are an emerging field of biotechnology research and their applications are only just beginning to be documented in scientific publications and patents. US scientists and research organizations dominate this field in terms of the number of scientific papers, the most cited work, and the number of patent applications. Certainly, political initiatives taken by the US government, including the Bayh-Dole Act, have succeeded in encouraging universities to patent inventions from federally funded research. Clearly technology transfer arms of universities in the United States are far ahead of those in other parts of the world in filing patents on SNPs.
In terms of patenting and publishing, US research-based biotechnology companies are more active than any other group. Their domination is most apparent in those areas related to healthcare (including gene therapy, diagnostics, or predictive healthcare) or tools to implement such new techniques (including microarrays or microfluidics).
We found no evidence of patents or papers on this area from giant pharmaceutical firms such as Merck (Rahway, NJ), Novo Nordisk (Copenhagen, Denmark), or Aventis (Strasbourg, France). Of course, biotechnology companies are much more likely to patent their intellectual property than pharmaceutical companies, which tend to keep hold of their intellectual property as trade secrets. Thus, a lack of patent applications should not necessarily be viewed as an indication of a lack of interest in the area by the pharmaceutical sector; after all, the SNP consortium includes drug companies such as GlaxoSmithKline, AstraZeneca, Aventis, Bayer (Leverkusen, Germany, Bristol-Myers Squibb (Princeton, NJ), Novartis (Basel), Pfizer (New York, NY), Hoffman−La Roche (Nutley, NJ), and Searle (now part of Pharmacia, Skokie, IL).
Interestingly, four drug companies involved in the SNP consortium have elected not to devote efforts to publishing and filing patents in the area of SNPs: Bayer, Bristol-Myers Squibb, Pfizer, and Searle. This could be explained in several ways: by a lack of interest in investing money in controversial gene patents rather than chemical patents; by a decision to keep the information in house as a trade secret; or by an investment in generating publications or patents insufficiently large to be detectable in the specific databases we analyzed.
Our survey also reveals the activity of large companies (such as Motorola and Packard Bioscience, which is now Perkin Elmer) that specialize applying information technology to SNP research. This is because the tools of robotics, miniaturization, and information technology required for many other areas of genomics also apply to high-throughput SNP genotyping tools. Motorola (together with IBM; Armonk, NY) is collaborating with pharma companies and academic centers in the ongoing efforts of the international SNP consortium.
The presence of an established biotechnology industry in the United States and excellent research infrastructure has certainly facilitated the US lead in this area. Despite heavy investment in Europe to encourage the creation of new biotechnology startups, the data clearly show that European companies and researchers lag behind their US counterparts in SNP research. Considering the youth of the SNP field and the more encouraging recent environment for biotechnology in Europe, it is rather remarkable that the United States has established such a dominant position in this comparatively new area. Perhaps the prominent role of US centers in the human genome project also contributed to their lead in SNP research. Certainly, the meager number of European and Japanese firms in this field (especially compared with other areas of biotechnology) clearly indicates the poor competitiveness of Europe in this area of biotechnology.
The US biotechnology sector clearly leads in applying SNP research to healthcare and agriculture. This could be a consequence of many factors, including an established entrepreneurial culture, an excellent infrastructure for startups, world-leading expertise in SNP research, and a more favorable environment for patenting genes and gene variants at the national patenting agencies.
FURTHER READING Drabek, J. A commented dictionary of techniques for genotyping. Electrophoresis22, 1024–1045 (2001). | Article | PubMed | ISI | ChemPort |
Gut, I.G. Automation I genotyping of single nucleotide polymorphisms. Hum. Mutat.17, 475–492 (2001). | Article | PubMed | ISI | ChemPort |
De Looze, M.A., Coronini, R. & Joly, P.-B. A note on recent trends in knowledge creation and appropriation through genomics: a scientometric analysis. Int. J. Biotechnol.3, 1–2 (2001).
Saviotti, P.P., de Looze, M.A., Michelland, S. & Catherine, D. The changing marketplace of bio-informatics. Nat. Biotechnol.18, 1247–1249 (2000). | Article | PubMed | ISI | ChemPort |
The International SNP Map Working Group. A map of human genome sequence variation containing 1.42 million single nucleotide polymorphisms. Nature409, 928–933 (2001). | Article | PubMed | ISI | ChemPort |
1.8 million SNPs in the public arena.
