Box 1. Methodology

The fate of individual patent applications is often difficult to determine because the granting process is often lengthy, and applications have not always been made public. As patent offices have different propensities to divide submitted inventions into two or more patent applications, our analysis is based on patent families as well as individual patents. A patent family comprises all the patent applications and granted patents resulting from a specific invention. The actual number of patent applications uncovered by our study is likely to be an underestimate because many applications are abandoned at the provisional stage, and data on USPTO applications submitted before November 29, 2000, but not yet granted, are unavailable. Our analysis used Thomson Scientific's GENESEQ and World Patent Index databases to identify patent families claiming human DNA and/or other nucleic acid sequences that were published between 1980–2003. Other data (e.g., legal status of granted patents) was obtained from the USPTO and EPO online databases. Data collection for JPO and USPTO patents was completed by June 2005, and for EPO patents by September 2005. Subsequently the study was augmented by interviews with individuals from ten biotechnology firms, ten large pharmaceutical firms and ten public sector organizations (that is, charities, government research organizations and universities) in the EU, the US and Japan. The majority of interviewed organizations are in the top 50 most active assignees in the field. Collectively those interviewed hold over 29% of the patent families in our data set.