Should companies set up databases in Europe?

Companies that are in the biological data business should think about establishing databases in the European Union (EU). Many EU countries have implemented laws following the 1996 directive (96/9/EC) on the legal protection of databases (“Database Directive”). These give database builders specific protection—“database rights”—recognizing the work and costs entailed in compiling, verifying, and presenting data.

Legislation also affords protection to the structure of databases. Database producers rely on both technical means and on legal devices, including patents, trade secrecy law, unfair competition and contractual arrangements, to secure conditional access to their databases. However, the EU directive gives another level of protection based on the premise that databases are expensive to create, but cheap to copy. There is no comparable legal protection for databases in the United States yet. Copyright protection in the United States has to involve “creative originality” and, as US law stands¹, building databases is not creative enough.

The “database right” in a DNA database would almost undoubtedly represent valuable intellectual property, especially with the uncertainty that still surrounds DNA sequence patenting. A database builder does not have to address issues such as discovery-versus-invention, novelty, inventive step, industrial application, or “ordre public”. Also, since the “database right” arises by operation of law, there is no need to register a database or to discuss its accrual or scope with a patent office. The “database right” allows its holder to prevent extraction or reutilizaton of substantial parts of the database for 15 years after completion of a database. In fact, the right may pertain in perpetuity given that it can be rolled over following any substantial modification to the database that requires a substantial investment. Such modifications include the routine business for builders of DNA databases: extensions, deletions, and amendments.

Although the value of database rights is clear, it is less certain that databases of DNA sequences qualify for them. The first thing to be said is that the directive was not designed with this type of database in mind. Nothing in the directive says, explicitly, that DNA databases are covered. Equally, nothing explicitly excludes them either.

To trigger the protection of the directive, a DNA database must fit within its definition of “database”; that is, it must be “a collection of works, data or other independent material, arranged in a systematic way and individually accessible.” At first glance, DNA sequence databases seem to qualify. Clearly, there are elements of “collection,” “systematic arrangement,” and “individual accessibility.”

From a legal point of view, it can be argued that sequence data or at least the annotations in a sequence record, do qualify as “works,” even though the term generally means copyrightable works, in the sense of literary, artistic, or other creations. Such argument may, however, be superfluous since DNA sequences qualify simply as “data.” Though it has been argued that “data discovered in nature” cannot be the subject of a Database Right², the directive describes “data” as “figures,” “facts,” and “data,” and does not limit the nature of the data. Thus DNA sequences or in any event cDNA sequences should qualify.

It is, however, still necessary that DNA sequences are shown to be “independent material.” The purpose of the requirement for independence is to exclude from database protection works such as books; the constituent chapters of a book are not independent. When DNA sequences are partial and their “meanings” or functions unknown, they may be considered independent. However, given that genomes have been described as “books of life,” might it not be argued that sequences with a known function are “chapters” or “paragraphs” of an entire genome? The issue of independence of the data is largely unresolved. It is at least an interesting possibility that a DNA sequence database is less readily protected the more complete it becomes.

Another interesting nuance of the Database Directive is that a database right will only attach to a maker of a database who can demonstrate, quantitatively or qualitatively, a substantial investment¹. It has been argued that no database right would arise if a database was a mere spinoff of the actual object of the investment³. That can hardly be said of the DNA sequence databases currently being built: both the publicly funded bodies and the genomics companies have been “building up DNA sequence databases” as their stated missions and apply most of their resources to that end. The same might not apply as readily to, for instance, databases of human genetic variation that emerged as part of broad health surveys.

A major complication in considering database rights for DNA sequence databases is the fact that they are built up from the contributions of numerous compilations of partial sequences. How can one decide who built a DNA database and who, consequently, owns the database rights? Even databases created by commercial genomics companies are combinations and extensions of a mix of commercial and nonprofit databases. Each of the researchers who have submitted new data to GenBank, for example, could claim to have invested in the collection of the database, as could others who keep the database comprehensive, current, and accurate, or annotate it⁴.

Legally speaking, the directive covers this issue. It says that the maker of a database (and thus the holder of the database right) is the person who takes the initiative to make the investments and bears the risks. It explicitly excludes subcontractors, a status that might apply inter alia to annotation staff. But external publicly funded research teams submitting multiple sequences or even genomes might not consider themselves subcontractors and might, therefore, wish to retain a database right with respect to their submission. Jointly held database rights are certainly a legal possibility².

The commercial reality, however, is more complicated. Some contributors may waive the joint rights they could enjoy in DNA databases, but they are under no obligation to do so. Indeed, they may be required under the terms of their grant or tenure to exercise those rights. And if they did, database makers would have to negotiate numerous contracts licensing or swapping data⁵. Licensing is unlikely to be straightforward because it is not compulsory and would be complicated further by the fact that the Database Right will be denied to contributors outside the EU.

Both public and private contributors to DNA databases could enjoy a Database Right. Given the number of biological databases and their architecture, a proliferation of fragmented and overlapping rights could emerge. The detailed arrangements for sharing the rights could be prohibitive to further research and development, unless we can devise a Database Right pool with an accompanying clearance system.