Introduction
Can personal genomic sequence data be protected by IP rights?
Datasets
Patents
Trade secrets
Intellectual property from publicly funded research
Comment


Introduction

This article is part of a series that provides a comprehensive and integrated analysis of the various legal dimensions applicable to the most basic kind of human genomic data, "personal genomic sequence data" – that is, the genomic sequence data (often referred to as the "raw" genomic data) of a particular person (in contradistinction from aggregated genomic sequence data of multiple persons).(1) In particular, this article focuses on the IP rights dimension of personal genomic sequence data.

Can personal genomic sequence data be protected by IP rights?

"Intellectual property" refers to creations of the mind, such as inventions, literary, artistic and musical works, designs, and symbols, names and images used in commerce. Personal genomic sequence data per se – also referred to as "raw" data – is not a creation of the mind, and therefore does not attract IP rights. However, personal genomic sequence data can be a useful building block to create intellectual property. This article considers three ways in which personal genomic sequence data is used:

  • to compile genomic datasets;
  • to develop new technologies; and
  • as trade secrets.

Datasets

In South Africa, if a person creates a compilation of personal genomic sequence data (ie, a genomic dataset), copyright in such a dataset will automatically vest in such a person – as per the Copyright Act 98 of 1978. Importantly, copyright vests in the dataset as a whole, and not in the constituent personal genomic sequence data.

This raises the question of what the effect would be, in property law, of including an instance of personal genomic sequence data in a dataset. In principle, it would be possible to sell or rent out the use of the dataset, in addition to assigning or licensing out the copyright in the dataset. But does this mean that the personal genomic sequence data would lose its status as a legal object and become a mere part of the dataset? This would only be the case if personal genomic sequence data becomes permanently part of a dataset; becoming part of the legal object that is the dataset by way of accession and thereby losing its status as an individual legal object. However, this is not so. The data of a particular constituent personal genome can, with relative ease, be separated from the dataset. Accordingly, it does not accede to the dataset.

This means that if compilers of genomic datasets do not acquire ownership of all the constituent personal genomic sequence data, they need to obtain the consent of all the owners of the constituent personal genomic sequence data to be able to out-license use of the dataset (or perform any other act that would affect the ownership rights of the owners of the constituent personal genomic sequence data). It would clearly simplify matters if a dataset were created by the same research institution that owns the constituent personal genomic sequence data. However, if multiple research institutions are involved in the provision of the constituent personal genomic sequence data and the creation of the dataset, a data management plan should be put in place to provide for issues such as ownership of the constituent data and how the dataset is to be used.

It is also important to note that the creation of a dataset of personal genomic sequence data and the consequent vesting of copyright in it also does not diminish the privacy rights of the data subjects whose genomic sequence data comprise the dataset. If using the dataset falls within one of the research exceptions of the Protection of Personal Information Act 2013 (POPIA), consent from the data subject need not be obtained to use (or out-license use of) the dataset. However, if use of the dataset is to be out-licensed to someone beyond the borders of South Africa, POPIA sets additional requirements that must be complied with.(2)

Patents

In the present context, a patentable invention may be in the form of a diagnostic tool, a therapeutic composition or, since South African courts are likely to take into account article 5 of the EU Biotechnology Directive,(3) one or more isolated gene sequences – provided they have a demonstrated functional use.

Importantly, in order for an inventor to patent an invention, the inventor need not own the object to which the invention relates. Therefore, an inventor need not own a personal genomic sequence data fragment – a gene sequence – in order to patent such a gene sequence. Similarly, the fact that the object that the invention relates to is also the object of privacy rights does not necessarily impede the invention's patentability. Patents do not bestow a right to use the invention. In order to use a patented gene sequence, permission would be required from both the owner of such data and the data subject.

The South African Patents Act 57 of 1978 provides that a patent may not be granted for an invention where its publication or exploitation would generally be expected to encourage offensive or immoral behaviour. However, this provision has never been tested in the courts. It may be understood that the fact that third parties have property or privacy rights in relation to the object of a patent application is not per se offensive or immoral.

What would an offensive or immoral scenario look like in the genomics context? Consider the following hypothetical scenario: an inventor applies for a patent on a diagnostic method, or a method of selecting a medical compound or composition for use in the treatment of a particular individual or ethnic group. The invention relates to a gene sequence obtained from an individual or persons belonging to such an ethnic group. However, the inventor intends to commercialise the patent in such a way as to make it unreasonably expensive to access, or the inventor intends not to commercialise the patent at all and only use it to block potential competitors. In this scenario, it is likely that the Patents Act's offensive or immoral provision would be triggered. Ergo, patent rights relating to genomic data should be used in a manner that is not only respectful of individuals' privacy rights and ownership rights, but also for purposes that are aligned with the public interest.

Trade secrets

In keeping with the Agreement on Trade-Related Aspects of Intellectual Property Rights (1994), the common law provides the general requirements for confidential information to qualify as a trade secret – namely, that such information must:

  • be capable of use in trade or industry;
  • have economic value to the proprietor;
  • not be public knowledge; and
  • be maintained by the proprietor (trade secrets are not owned but possessed) as a secret.

If these requirements are met with regard to personal genomic sequence data, such data may qualify as trade secrets. Accordingly, assuming that owners of personal genomic sequence data are also in possession of such data, if they keep the data secret (eg, by using non-disclosure agreements with employees and research partners), they will enjoy trade secret protection against misappropriation or unauthorised use.

Given the fungible nature of personal genomic sequence data, the question arises as to whether an individual's personal genomic sequence data can ever be maintained as secret by a research institution, as the data subject can conceivably provide biological samples to any number of other research institutes for genomic sequencing. Data subjects can even, in theory, decide to publish their genomic data online – open for anyone to use. These possible eventualities are of a practical nature and do not, in principle, disqualify personal genomic sequence data as trade secrets. The requirement is that the relevant personal genomic sequence data must, as a matter of fact, not be public knowledge.

These possible eventualities illustrate that trade secret protection does not limit the rights of the data subject – instead, it is the other way round: the rights of the data subject can destroy a trade secret. Accordingly, where a research institution is the owner of personal genomic sequence data, trade secret protection can add a useful weapon to its arsenal of legal remedies that can be used against other research institutions, but not against the data subject. The benefit of trade secret protection comes at the cost of a self-imposed restriction on ownership – namely, keeping the personal genomic sequence data confidential.

As with patentable inventions, it is important to remember that the proprietor's right to use or commercialise personal genomic sequence data is subject to the data subject's privacy rights, as well as statutory limitations on use and commercialisation, such as those provided in the National Health Act 61 of 2003.

Intellectual property from publicly funded research

No discussion of IP rights in South Africa is complete without mentioning that intellectual property which emanates from research that received financial support from the South African government – no matter how small – is subject to an additional layer of regulation, in terms of the South African Intellectual Property Rights from Publicly Financed Research and Development Act 51 of 2008. The National Intellectual Property Management Office exercises a wide range of statutory powers, and is mandated to ensure that such intellectual property is "protected, utilized and commercialized for the benefit of the people of the Republic". While this mandate restricts the rights of the owners of affected IP rights, it does not restrict the ownership rights of personal genomic sequence data owners or data subjects whose personal genomic sequence data is contained in genomic datasets.

Comment

The best metaphor to describe the multidimensional legal nature of personal genomic sequence data is a bundle of rights applicable to personal genomic sequence data. This bundle of rights may include common law ownership, personality rights and IP rights. The interplay between these various rights will determine the measure of control over the personal genomic sequence data that different stakeholders, such as research institutions, research participants, research funders, and any downstream entities that wish to commercialise innovations derived from or encompassing personal genomic sequence data, can lawfully exercise at various steps in the innovation process.

The practical consequence of the bundle of rights applicable to personal genomic sequence data is that public policies and institutional policies should acknowledge and deal with personal genomic sequence data in all of these dimensions. In other words, a policy that deals with personal genomic sequence data only in one or two dimensions leaves a lacuna that can lead to legal uncertainty or undesirable consequences. For example, consider a scenario where a research institution collects biological samples from human research participants, extracts the DNA, and sequences the genomes. However, the research institution has no agreements in place with the research participants regarding ownership of the personal genomic sequence data. Furthermore, the research institution labours under the misapprehension that personal genomic sequence data is not susceptible of ownership, and therefore never forms the intent to acquire ownership. The result is that personal genomic sequence data is generated, but nobody acquires ownership in such data – at least not at the time that the data are generated. Personal genomic sequence data are res nullius—something that belongs to no one.

This is problematic because the first person who intends to become the owner of a res nullius and takes control of it acquires ownership in it. To expand on the scenario described above: what if the research institution shares the personal genomic sequence data with its international consortium partners, but still fails to address ownership of the data in the consortium agreement? At that stage, any one of the consortium partners can lawfully acquire ownership in the data – at least their copy of it – simply by willing it and being in effective control of their copy of the data. Alternatively, the research institution compiles a dataset and out-licenses use of the dataset to a foreign company. Again, if the licence agreement does not prohibit the licensee from claiming common law ownership in the data in the dataset, the licensee can easily and lawfully acquire ownership in the data. This may have drastic, unintended consequences for the research institution that initially generated the data, but which neglected to properly deal with the ownership dimension of personal genomic sequence data. This underscores the importance of dealing with all of the legal dimensions of personal genomic sequence data – as notably highlighted in this series of articles.

For further information on this topic please contact Joanne van Harmelen at ENSafrica by telephone (+27 21 410 2500) or email ([email protected]). The ENSafrica website can be accessed at www.ensafrica.com.

Endnotes

(1) This series on the legal nature of personal genomic sequence data is a summarised version, containing excerpts, of "The multidimensional legal nature of personal genomic sequence data: A South African perspective" by Thaldar et al (2022). For earlier articles in this series, see:

(2) Townsend, B A (2022), "The lawful sharing of health research data in South Africa and beyond", Inf Commun Technol Law 31 (1), 17–34.

(3) Directive 98/44/EC of the European Parliament and of the Council on the legal protection of biotechnological inventions, 1998.