Technological advances in science often involves being able to improve the way a task is carried out, whether the task can be done quicker, more conveniently and/or with more accuracy. In certain specialties of science, it is desirable to find ways to automate a task or to collate data from several sources to improve efficiency of data analysis, which is often the case with bioinformatics, a highly specialized interdisciplinary field that combines the intricacies of biology and genetics with computer science, mathematics, and statistics.
In the era of personalized medicine, bioinformatics has, and will continue to play, an essential role in advancing science and medicine. For example, mutational analyses conducted on nucleic acid sequences, genes or proteins, have helped scientists to better understand the impact of any variations to the natural sequence of these molecular components; phylogenetic analysis has helped developmental biologists to decipher and trace evolutionary lineages in a variety of organisms; and high-throughput analyses has assisted with the collection, processing and systematic analysis of raw data to produce meaningful results that can be utilized by scientists for a number of downstream applications.
Subject matter eligibility of bioinformatic inventions in key jurisdictions
In an earlier article the patentability of computer implemented inventions has been discussed. In brief, the decision of RPL Central (Commissioner of Patents v RPL Central Ltd [2015] FCAFC 177 (11 December 2015) and Research Affiliates LLC v Commissioner of Patents [2014] FCAFC 150 (10 November 2014) held that “mere” business methods or schemes are not patentable subject matter. However, computer implemented inventions that solve a technical problem, whether within a computer itself or in a technical field, are patentable subject matter.
Along similar lines, inventions that use software are patentable in Europe, as long as there is a technical purpose for that invention outside the software itself. This means that in Europe, it is possible to re-structure a claim directed towards the abstract method to one that focuses on the technical application of that method, since the latter is deemed patentable. Often the technical purpose in bioinformatics relates to the biological purpose for which the bioinformatics method applies. Therefore, a claim directed towards the unique features or algorithms in a bioinformatics method alone is unlikely to satisfy the requirements for patentable subject matter, whereas a method for determining the evolutionary lineage of a species using a bioinformatics method, or more commonly, a method for diagnosing a particular disease in a patient sample using a bioinformatics method, are both likely patentable in Europe.
With respect to inventive step, such bioinformatics methods are normally assessed in 2 parts, the first of which involves consideration of the technical features of the invention, which may include consideration for the technical equipment or physical steps that are used when carrying out a method. Secondly it also involves consideration for the non-technical (abstract) features of the bioinformatics method which are capable of contributing towards the technical effect, and thus the technical purpose, of the described invention. This means that any non-technical feature that is peripheral to the technical purpose of the invention, for example graphical user interfaces, are not relevant to an assessment of inventive step.
In the UK, similar to Europe, unless a bioinformatics method has a technical contribution towards the functionality of an invention, the abstract method alone will be unlikely to be accepted by the UK IPO. Seeking patent protection for bioinformatic methods in the UK differs slightly from the process of obtaining patent protection in Europe, since there have been a few occasions whereby AI invention have been considered patentable in the UK. For example, a method of determining the percentage of blood being pumped from the heart, based on training a neural network with ultrasound imaging data, has been deemed eligible subject matter for a patent in the UK. As a result, there may be scope in achieving patent protection for some “excluded matters”, such as methods of medical treatment or methods of diagnosis on humans or animals, although this remains a developing area of law.
Challenges also exist in the US with respect to seeking patent protection in the field of bioinformatics. For example, patents directed towards haplotype phasing have been considered ineligible patent subject matter (re Board of Trustees of Leland Stanford Junior University, Case: 20-1012, 11 March 2021).
However, in good timing, the USPTO has recently announced its Deferred Subject Matter Eligibility Response (DSMER) Pilot Program, which will only be available to US patent Applicants by invitation. Briefly the program is designed to evaluate how deferred Applicant responses to subject matter eligibility objections affect examination efficiency and patent quality, when compared with the standard US prosecution practice. How the program works is that it provides the Applicant with a limited waiver in the US, with respect to any subject matter eligibility objections raised in an Office Action. Whilst a reply is still required to be filed in response to any Office Action issued to the US patent application in question, the limited waiver allows the Applicant to defer any arguments and/or amendments in relation to the subject matter eligibility objections, until either the Final Office Action is issued, or all other remaining objections have been addressed or withdrawn, whichever is earlier. Aside from the benefit of deferring any subject matter eligibility arguments and/or amendments, the relevant US patent application continues to undergo standard examination procedures. It is anticipated that further updates on this program will likely be made later this year.
Artificial intelligence and bioinformatics
In recent times, there has been an increase in the number of patent applications filed in relation to AI inventions. In particular, the Japanese Patent Office has seen an increase in the number of filings for AI-related inventions, most particularly on inventions based on the concept of deep reinforcement learning; an upward trend that appears to be continuously growing in light of technological advances. Such AI-related inventions include advancements made in the fields of Medical Diagnosis, Image Processing, Health Care and other sectors.
Often in healthcare there are scenarios whereby decision-making, or even clinical sample screening is performed by AI to produce results that are just as good, or even superior to that produced by a human operator. It is anticipated that AI will have an increasingly important role in carrying out routine tasks in research, whether this be for identifying protein sequences, identifying interactions between target drugs, or even in the analysis of nucleic acid sequences.
Consequently, there may be occasions when AI makes new discoveries during the process of performing routine research tasks. We have recently published an article discussing Australia’s stance on whether artificial intelligence can be considered an inventor under the Patents Act 1990 (Cth) (https://www.mbip.com.au/the-future-is-now-dabus-update/). In brief, there are a number of jurisdictions in which an inventor must be a human being in order to have entitlement to a patent. Such jurisdictions include the Australia, New Zealand, the US, the UK, Germany, Korea, amongst others. However, there has been some notable exceptions, in that South Africa granted the world’s first patent to an AI inventor, the DABUS system, back in 2021.
Advice for applicants seeking to protect bioinformatics inventions
In general, it is advisable that any patent application seeking protection for a bioinformatics method should be made by person and not an AI system. It should include a full disclosure of the computational algorithms used in the method, and provide a clear disclosure on how the said algorithms contribute towards the technical purpose of the invention. Care should also be taken as to the wording used in the specification, particularly in the abstract, claims, and title, so as to ensure that Examiners are clear on the technical field of the invention. It is recommended that a full disclosure of the above should be provided at the time of filing, given that some countries such as Europe do not permit post-filing amendments, which could be deemed as added subject matter.
Bioinformatics is an important scientific area that is likely to grow alongside developments in personalized medicine. Given that bioinformatics is such a multi-faceted complex area of bioscience, patenting in this space requires the assistance of a multi-disciplinary team that is equipped with the appropriate technical skills that combine the knowledge of the life sciences with that of computer science, in order to navigate the challenges associated with seeking patent protection in this area. The team at MBIP have the necessary expertise to assist clients with this.