Scientific linkage of science research and technology development: a case of genetic engineering research

Transcription

1 Scientometrics DOI /s Scientific linkage of science research and technology development: a case of genetic engineering research Szu-chia S. Lo Received: 21 August 2008 Ó Akadémiai Kiadó, Budapest, Hungary 2009 Abstract In this study, the author tried to demonstrate the linkage between science research and technology development through non-patent citation analysis to reveal that the important knowledge resources from science research had significant impact on technology development. Genetic engineering technology was the field examined in this study. From the references listed in the patents, it was observed that the technology development in genetic engineering was influenced heavily by the research done by public sector. Over 90% of the citations were non-patent literatures, and the majority of nonpatent citations were journal articles. Citing preferences, such as country preference and institute preference were observed from the data included in this study. Keywords Scientific linkage Non-patent citation Basic science research Technology development Introduction It is widely accepted that science research is the driving force behind technology development and economic growth both in scientific and economic communities. The utilization of science research output in the development of high technology is seen as presentation of the usefulness of science. It is helpful to the researcher to understand the dissemination of knowledge gained from science research to technology development by constructing the linkage between science research and technology development. By the same token, the demonstration of the usefulness of science is valuable to the decision maker during the process of drafting strategic plan. During the past several decades, plenty studies were done to show the productivities and research impact. There are quite an amount of studies which applied the methods adopted from bibliometrics while periodical articles were used for analyzing. Patent information becomes more accessible in recent years and more visible in relevant studies, such as the work S. S. Lo (&) Graduate Institute of Library and Information Science, National Chung Hsing University, Taichung, Taiwan, ROC losu@dragon.nchu.edu.tw

2 S. Scarlett Lo done by Narin (1995). In some cases, journal articles were treated as representations of the results of science research and patents were seen as the output of technology in others. The publication count for both types of materials was used to show the research productivities, and the citation count was done for the analysis of research impact. In early studies, the researches focus on the discussion of whether there was any correlation between science and technology (Collins and Wyatt 1988; Narin and Noma 1985; Narin and Olivastro 1992). Empirical studies were carried out later on to enforce the validity of the linkage and the importance to establish the linkage (Anderson et al. 1996; Narin et al. 1997), and several researches showed the dependence of industry development on public science (Narin et al. 1997; McMillan et al. 2000). Citation analysis was the most common approach taken to define the correlation between science and technology (Bhattacharya et al. 2003; Collins and Wyatt 1988; Meyer 2000a, 2000b; Narin 1994; Verspagen 1999). Peer opinion regarding the transformation from science to technology was drawn by conducting interviews (Fagerberg 1987, 1994; Rosenberg 1982). In this study, the author tries to demonstrate the linkage between science research and technology development, highlighting the impact of science research on the technology development through non-patent citation analysis. The study aims to reveal the important entities of science research that had significant impact on technology development, and if there was any citing preferences existed. Genetic engineering was the chosen subject for this study. Research problems The purpose of this study was to reveal the linkage between science research and technology development by examining the non-patent literatures cited by patents. Non-patent literatures, especially journal articles, were seen as representation of the output of science research and patents were taken as tokens for results of technology development. The author examined the patents granted to the primary organizations in genetic engineering research, which were identified by the results of patent counting on the patents granted by United States Patent and Trademark Office (USPTO) during 25 years ( ), as well as the patents and non-patent literatures cited by the identified patents. The linkage between citing patents and cited materials was reviewed to construct the link between science research and technology to answer the following questions about genetic engineering research. Did science research have impact on technology development? What were the sources of impact? If institutions from public sector showed greater influence on the research than the ones from private sector did? Was there any citing preference existed? Methods and data Methods This study took bibliometrics approach; Patent Count and Citation Count were used to show the research productivity and impact. Further analyses were done by employing

3 Scientific linkage of science research and technology development Fig. 1 Data sets included and analysis applied Bradford s model (Narin and Moll 1977; Garfield 1980) to identify the core assignees, and non-patent citations cited by the patents granted to the core assignees were examined in this study to reveal the linkage of the science research and technology development. The following attributes were examined in this study: types of the non-patent citations, the citation ages, the major resources of citations, and the key institutions contributed to the development of genetic engineering technology. Data The data source used in this study was USPTO Patent database. The study analyzed the 6,274 patents granted during the period of which were identified as genetic engineering patents by the International Patent Classification (IPC) and United States Patent Classification (USPC) numbers. The patents, which had the primary IPC numbers of Mutation or genetic engineering (C12N 15/00) and USPC numbers listed under the class Process of mutation, cell fusion, or genetic modification (435/440), were defined as basic genetic engineering patents. The core assignees were identified by taking the Bradford s model approach. The top 16 assignees, each was granted at least 50 patents, were located in the core zone of the Bradford s distribution. This study further examined 1,412 patents granted to the top 16 assignees and works cited by those patents, which included 4,001 cited patents and 35,447 non-patent citations. Figure 1 shows the workflow of this study, including the data sets created in this study and analyses done with the data. Findings Basic analysis Basic patent count and Bradford s model analysis were applied for productivity analysis. The results were presented by time, technologies, countries and assignees. Time distribution The study included patents issued during the period of 25 years, in which four time zones could be drawn: , , , and The number of patents granted increased dramatically during the period from 1996 to Even when the

4 S. Scarlett Lo Fig. 2 Annual count: number of patents issued declined after year 2000, it was believed that the research focus had been switched from basic research of genetic engineering to other applied field, such as pharmaceutical industry. Figure 2 shows the numbers of patents granted from 1980 to 2004 by year. Technologies distribution During further examination of the primary International Patent Numbers of the 6,274 patents, it was found that significant amount of patents involved with Recombinant DNA technologies. Among the 6,274 patents, 4,562 (72.71%) patents were dealing with the techniques of DNA recombination, including general process of DNA recombination, process of isolation, modifying DNA fragments and introducing foreign genetic materials. Comparing to the DNA recombinant techniques, there was only very limited number of patents granted that involved cell fusion and mutation by non-insertion foreign genetic materials methods. Besides the general process of the techniques mentioned above, close to 25% of the patents (1,527 in 6,274) involved application of genetic engineering process in preparation of compounds, proteins, enzymes, peptides and breeding species. Figure 3 shows the distribution of the patent techniques. Looking into the recombinant DNA techniques and comparing the numbers of patents granted in sub-domains, it was found that more effort and outcomes had been put into the research of Modifying DNA/RNA fragments and Introducing foreign genetic materials by using vectors. There were 2,215 patents granted in DNA/RNA fragment modification and 1,751 patents in introducing foreign genetic materials with vectors. By examining the distribution of patent technologies from the time point of view, it was observed that there was a technological shift during late 1990s and early 2000s. More patents were granted in DNA/RNA fragment modification during early stage and more patents were granted in Introducing genetic material after late 1990s. Productive countries With the domestic advantage, it was not surprising that United States was granted the most number of patents in genetic engineering research among the countries that applied US- TPO patents. The local advantage could be part of the reason, but not the only grounds for

5 Scientific linkage of science research and technology development Fig. 3 Distribution of patent technologies Table 1 Productive countries ([100 granted patents) Country Total United States ,286 1,376 4,559 Japan Germany France Great Britain Canada Netherlands the high productivity. The fact of being the founder of genetic engineering should not be over looked. 71% (4,559) of the patents were granted to the assignees based in United States. With 395 patents, Japan was one of the leading countries in genetic engineering research besides United States. Germany, France and Great Britain, European countries, were the other three leading countries, holding patent rights to 251, 224 and 210 patents. Table 1 shows the results of patents in countries that were granted over 100 patents in Genetic Engineering during the period of Productive assignees There were 1,300 assignees identified from 6,247 patents and 16 core assignees were singled out through Bradford analysis. All the assignees were ranked by the number of patents granted and divided into three zones, with each zone contributed one-third of patents. The assignees listed in the first zone, also called core zone, were identified as productive assignees. Among the productive assignees, two California based institutes,

6 S. Scarlett Lo Table 2 Productive assignees, top 16 ([50 granted patents) Assignee No Co St Tp University of California (Berkeley) 181 US CA EDU INCYTE Pharmaceuticals 127 US CA COM SmithKline Beecham Corporation UK/US PA COM Dept of Health and Human Services, US 99 US DC GOV Pioneer Hi-Bred International, Inc. 98 US IA COM Genentech, Inc. 95 US CA COM Monsanto Company 85 US MO COM The General Hospital Corporation 81 US MA COM/ORG Human Genome Sciences 79 US MD COM Chiron Corporation 74 US CA COM Harvard University 68 US MA EDU The Johns Hopkins University 60 US MD EDU Eli Lilly and Company 58 US IN COM Merck & Co., Inc. 56 US NJ COM The University of Texas (Austin, TX) 53 US TX EDU Genetics Institute, Inc. 53 US MA COM University of California (Berkeley) and INCYTE Pharmaceuticals, 1 and SmithKline Beecham, 2 were the top three assignees that were granted the most number of patents, 181, 127 and. Table 2 lists the numbers of patents granted to the core assignees. With the advantage of being near by University of California (Berkeley), the leading institution for genetic engineering research, it was found that several assignees based in the bay area from private sector also showed productive strength in this field. Besides granted more patents in Modifying DNA or RNA fragments, all the top three productive assignees specialized in various sub-domains of genetics engineering research. University of California (Berkeley) was granted significant amount of patents in Introducing genetic materials, SmithKline showed gaining in General process of DNA recombinant, and INCYTE Pharmaceuticals was granted more patents in Preparation of protein and enzyme. Science linkage analysis The cited references listed on the patent front page showed that genetic engineering research was highly dependent on prior art, especially the researches done by public sector. Not only patents were cited, significant amount of non-patent literatures also contributed to technology development. Examining the patents granted to the top productive assignees, there were 4,001 patents and 35,447 non-patent literatures cited by 1,412 patents. Each patent listed references in average, in which, one tenth were patents and the others were non-patent literatures. Variation of the citation behaviour was observed among the 1 INCYTE Pharmaceuticals was founded in The name was first changed to INCYTE Genome and changed again to INCYTE Corporation in INCYTE Corporation was moved to Willington, Delaware in 2004 and shifted the research focus to drug research. 2 SmithKline Beecham was merged with Glaxo to form GlaxoSmithKline in 2001.

7 Scientific linkage of science research and technology development Fig. 4 Distribution of patents and non-patent citations of top 16 productive assignees productive assignees. In average, the patents granted to General Hospital (Boston), Harvard University and Chiron cited up to 50 references each, while the ones granted to University California (Berkeley) and Genetic Institutes cited relatively less, with averagely 15 references per patent. It could imply the assignees in the former group focus on the value added works and the latter ones were more adventurous in genetic engineering research. The ratio of cited patents and non-patent literatures shows that all assignees cited more non-patent literatures than patents without exception. More than 90% of citations listed in the patents granted to INCYTE Pharmaceuticals, General Hospital (Boston), Harvard University, Human Genome, John Hopkins University and Genetic Institute were non-patent literatures. Even if some of the assignees did cite more patents, the percentage of cited patents was still less than 20% of total citations. Figure 4 shows the distribution of the cited patents and non-patents literatures. Time distribution The author further checked the numbers of patent citations and non-patent citations by years of patents issued to see if there were any changes in citation behaviour during different time periods. The results showed that non-patent citations dominated the research impact during all periods except 1986, when the number of patent citations were greater than the number of non-patent citations. It also showed that the numbers of non-patent citations increased over the years. The linkage of science research and technology development was getting more intense as the genetic engineering research progressed. Figure 5 shows the distribution of annual patent and non-patent citations. Material types of non-patent citations Among all types of non-patent citations, information extracted from monographs, proceeding papers, samples from Gene Bank and journal articles were the foundations of genetic engineering technologies. If the journal articles were seen as tokens of the output of

8 S. Scarlett Lo Fig. 5 Distribution of patents and non-patent citations time analysis Fig. 6 Material types of non-patent citations basic science research and the patents were taken as presentations of results for technology developments, it could be addressed that genetic engineering research depends heavily on the research of basic science. It was found that journal articles took a major part in nonpatent citations, accounting for 28,693 (89%) in total non-patent citations. A few things to be aware of are that, some patents were listed in the section of non-patent literatures, and the material types of 110 citations could not be identified due to lack of information. Figure 6 presents the types of non-patent citations and the distribution of citations in each material type. Looking into the sources of cited journal articles, 1,521 journal titles were identified. Among them, title of Proceedings of National Academic Science was cited more frequently

9 Scientific linkage of science research and technology development Table 3 Top 10 highly cited journals Journal title Subject Times cited Proceedings of the National Academy for Science Multi-science 2,929 Nature Multi-science 1,923 Science Multi-science 1,919 Journal of Biological Chemistry Biochemistry molecular 1,794 Cell Biochemistry molecular 1,489 Journal of Virology Virology 762 Nucleic Acids Research Biochemistry molecular 621 EMBO Journal Biochemistry molecular 604 Molecular and Cellular Biology Biochemistry molecular 579 Gene Genetics 503 Table 4 Top 10 author affiliations of non-patent citations Organization Count Harvard Medical School 121 National Institutes of Health 69 Genentech, Inc. 67 Johns Hopkins University 54 Massachusetts General Hospital 50 Stanford University School of Medicine 46 University of California, Berkeley 42 University of California, San Diego 41 Washington University School of Medicine 35 Massachusetts Institute of Technology 31 than others, with 2,929 times identified. Apart from Proceedings of National Academic Science, Nature, Science, Journal of Biological Chemistry and Cell were also highly cited. Table 3 lists the top 10 highly cited journals and the frequency of citation for each title. Origin of genetic engineering researches Further examination on first authors affiliations was done for samples of cited journal articles. One tenth (2,830) of the cited journal articles were checked and author information of 2,720 sampled articles could be allocated and obtained during the period of the study. Researches done in the United States were the core sources for the research impact. Among the 2,720 authors examined, there were 1,867 (68.64%) authors from United States based organizations. Comparing to the geographic distribution of assignees of cited patents, research impact from institutes based outside United States increased, such as the ones based in Great Britain, Japan, Germany and France. From the review of the authors and affiliations, the 2,720 authors were originated from 794 institutions. Among them, authors from Harvard University were cited the most, with 121 times in total. The authors affiliated with National Institute of Health and Genentech were also highly cited. Mapping the affiliations of cited patents and cited journal articles, it showed the institutional bias that

10 S. Scarlett Lo institutional self-citation was common practice by observation performed in the patents granted to Genentech. Similar citing behaviour was also seen in the patents granted to John Hopkins University. Table 4 lists the top 10 affiliations of the authors of cited journal articles. Referencing cycles The average citation age was 9.8 years. The mode value was 8 and the median value was 9. Although the result from related studies indicated that biotechnology might be one of the domains where there was no time lag or shorter time lag from science research to technological development and market value (Narin 1994), this study gave a different picture, in which, the citation age of 1,227 non-patent citations was over 20 years old by the time they were cited by the patents included in this study. Discussions The citation pattern and science linkage confirmed the impact of science research on the technology development in genetic engineering. From the references listed on the front page of patents, it was obvious that the technology development in genetic engineering was influenced heavily by the research done by public sector. Over 90% of the citations were non-patent literatures and over 89% of the non-patent citations were journal articles. Comparing to the findings of previous studies, industry development of genetic engineering is more reliant on public science. The United States still holds the domestic advantages. Not only the patents were granted to the United States based institutions, the sources of research impact also reflected the influence of the works done in United States based organizations. Although several assignees from private sector devoted resources into the science research and showed the impact through the citation linkage between patents and non-patent publications, the research works done by public sector, such as the ones by Harvard University, John Hopkins University and University of California, certainly demonstrated the strength not only in productivity but also in research impact. The information resources heavily used in science research were also highly regarded in technology developments. Highly public science dependent Comparing the distribution of cited patents and cited journal articles listed on the front page of patents granted to productive assignees, it was found that the number of cited journal articles exceeded the number of cited patents by significant margin. The difference indicated the link between technology development and science research of genetic engineering, and the link also represented the dependence. Country preference By examining the sample citations of patents granted to the productive assignees, it was found that the outcomes of genetic engineering research performed by the United States based institutions dominate the developments. Over 67% of the cited journal articles were written by authors affiliated with the United States based institutions.

11 Scientific linkage of science research and technology development Institutional preference Further examination of the authors affiliations in terms of institutional types shows that the patents granted to Universities tended to cite more works done by the authors who were also affiliated with universities. As for the private sector, the influence of public sector was also shown, but the distribution of sources of research impact was more balanced between public sector and private sector. It was also observed that the assignees from public sector had higher self-citation rate. No doubts that the development of genetic engineering research were highly dependent on prior art, especially the link to public science research. Continuing enquiries on citation behaviour could further reveal how the knowledge-transfer has taken place. Conclusions Genetic engineering patents (6,247) with cited references were examined in this study to establish the linkage between science research and technology development. The results showed that the development of genetic engineering technologies was heavily dependent on the science researches, especially the works done by public sector, in terms of numbers of journal articles cited in the selected patents and the origin of the cited works. It confirms the previous findings of Narin and McMillan s works (Narin et al. 1997; McMillan et al. 2000) and provides the evidence that the science research was the driving force behind technology development. From administration point of view, the linkage could be seen as a token to demonstrate the contribution of science research to industrial technology development. By calculating the frequency of citation, it offers a clear picture of main sources for research impact. Comparing the works cited in science research, the highly cited journals by patents were heavily used by both sides. Citing preferences were observed from the results. Works done by researchers in the same institutions or affiliations, as well as institutions geographically close to each other, tend to be cited more. For library professionals, those are useful information for policy making in library collection development. Genetic engineering was the only subject studied in this paper. Similar examinations need to be done in different technology areas to establish the commonality of the scientific linkage. References Anderson, J., Williams, N., Seemungal, D., Narin, F., & Olivastro, D. (1996). Human genetic technology: Exploring the links between science and innovation. Technology Analysis & Strategic Management, 8(2), Bhattacharya, S., Kretschmer, H., & Meyer, M. (2003). Characterizing intellectual spaces between science and technology. Scientometrics, 58(2), Collins, P., & Wyatt, S. (1988). Citations in patents to the basic research literature. Research Policy, 17(2), Fagerberg, J. (1987). A technology gap approach to why growth rates differ. Research Policy, 16, Fagerberg, J. (1994). Technology and international differences in growth rates. Journal of Economic Literature, 32(3), Garfield, F. (1980). Bradford s law and related statistical patterns. Current Contents, 19, Mcmillan, G. S., Narin, F., & Deeds, D. L. (2000). An analysis of the critical role of public science in innovation: The case of biotechnology. Research Policy, 29(1), 1 8.

12 S. Scarlett Lo Meyer, M. (2000a). Does Science push technology? Patents citing scientific literature. Research Policy, 29(3), Meyer, M. (2000b). Patent citations in a novel field of technology what can they tell about interactions between emerging communities of science and technology. Scientometrics, 48(2), Narin, F. (1994). Patent bibliometrics. Scientometrics, 30(1), Narin, F. (1995). Patents as indicators for the evaluation of industrial research output. Scientometrics, 34(3), Narin, F., Hamilton, K. S., & Olivastro, D. (1997). The increasing linkage between U.S. technology and public science. Research Policy, 26(3), Narin, F., & Moll, J. K. (1977). Bibliometrics. Annual Review of Information Science and Technology, 12, Narin, F., & Noma, E. (1985). Is technology becoming science? Scientometrics, 7(3 6), Narin, F., & Olivastro, D. (1992). Status report linkage between technology and science. Research Policy, 21(3), Rosenberg, N. (1982). Inside the black box: Technology and economics. Cambridge: Cambridge University Press. Verspagen, B. (1999). Large firms and knowledge flows in Dutch R&D system: A case study of Philips electronics. Technology Analysis and Strategic Management, 11,