Individual Patents

In this section, the citation networks for individual patents are examined. The patents of choice are those from Cohen and Boyer. The Cohen-Boyer technology for recombinant DNA is a path-breaking one by any standard, 'arguably the defining technology of modern molecular biology' (National Research Council, 1997, p. 40). There are in fact three Cohen and Boyer patents:

• 4740470 - Biologically functional molecular chimeras, 1988;

• 4468464 - Biologically functional molecular chimeras, 1984;

• 4237224 - Process for producing biologically functional molecular chimeras, 1980.

All three patents are continuations or continuations-in-part of patents originally filed in 1974, 19 76 and 1978. In simple terms, continuation patents are used when the patent examiner has filed the 'final rejection', allowing the applicant to alter the claims in various ways (see Grubb, 1999, ch. 6).

Fig. 10.8. Thirty-three subsectors networked with biotechnology excluding citations below a threshold of 2000.

The initial question arises as to which of these three is the dominant patent, the one implied when referring to the Cohen-Boyer patent, which has earned Stanford University hundreds of millions in royalties. To the non-specialist, a simple reading of the claims does not clarify the point.

Using network analysis for the three, the citation rates given in Table 10.1 were found.

From these data, it is apparent that Patent 4237224 is the dominant one of the three, and is of much use in biotechnology. We believe that network analysis is an effective methodology for identifying the key patent among several others. There are a number of instances when that might be desirable, such as determining which of the several patents is the core one to a proprietary drug.

Several other factors are also recognizable from these numbers. First, this is a very significant patent based on the large number of citations. Second, it has

Table 10.1. In-citations for three Cohen-Boyer patents.

Patent

Category

Subcategory

PTO class

Total

Direct

Indirect

Depth

Maximum width

4740470

3

33

435

70

4

66

5

34

4468464

3

33

435

788

58

730

7

62

4237224

3

33

435

14,244

200

14,044

12

93

In-components of companies 4740470 Cohen and Boyer (Standford University)

Corvas International, Inc.

Fig. 10.10. Entities in-licensing Patent 4740470.

longevity, as evidenced by the 'depth' or number of levels of citations. But perhaps what makes this patent so notable is the very large number of indirect citations to other patents, which subsequently cite Patent 423 7224. The relationship of direct to indirect citations can be better appreciated visually. Figure 10.9 displays the in-citations for Patent 4237224, but only for the first two levels - displaying all would make the figure far too cluttered to understand.

We posit, but leave for future verification, that including direct and indirect citations has greater explanatory power than direct citations only. Out-citations are important as well, with values for the three Cohen-Boyer patents shown in Table 10.2. Clearly, these patents are novel in the sense of not building

Table 10.2. In-citations for three Cohen-Boyer patents.

Maximum

Patent Category Subcategory PTO class Total Direct Indirect Depth width

701578

701578

265085 icos Corporation of America

to any degree on prior patented art. We posit, again leaving to future verification, that pioneering patents can be identified by the high ratio of in-citations to out-citations.

Other factors than the breakthrough aspect of a patent may, however, affect the aspect of in-citations. The Cohen-Boyer patents have been widely licensed on a non-exclusive basis by Stanford University (see Walsh et al., 2003, pp. 305 and 306), which might explain in part the large number of citations. In contrast, Merges and Nelson (1990, pp. 890-892) develop the case that the Wright brothers' unwillingness to license their broad technology airplane patents significantly delayed technical development in aircraft, whereas several patent scuffles operated similarly in the early days of radio technology. By comparing citation networks for openly licensed as well as restricted licensed patents it will be possible to get some insight into the effect on citations (and by inference technological advancement). The effects of open, non-exclusive licensing may perhaps be seen by examining the network of citing firms. The network of entities citing Patent 4740470 is shown in Fig. 10.10; the network of entities citing Patent 4237224 is visually attractive but so dense as to be uninterruptible.

0 0

Post a comment