We welcome back guest blogger and thought leader Professor Chris Holman at the University of Missouri-Kansas City Law School for a new two part series. Professor Holman authors the well-known Holman’s Biotech IP Blog and is the executive editor of Biotechnology Law Report.

DNA CopyrightIn a recent series of posts, I explored the idea of extending copyright protection to engineered DNA sequences.  Given GQ Life Sciences’ status as a leading provider of patent sequence search services, I thought it would be interesting to write a post considering what role sequence searching might play under a legal regime that recognizes copyright protection for engineered genetic sequences.

If copyright protection becomes available for a significant number of engineered genetic sequences, copyright sequence searching could become just as relevant as patent sequence searching.  For one thing, the Copyright Office might incorporate sequence searching into its copyright registration processes.  Users of pre-existing DNA sequences, including synthetic biologists engineering new genetic code, or firms involved in the manufacture, sale, and/or distribution of DNA and DNA-based products, might also be well advised to perform a copyright sequence search to ascertain, or at least confirm, the copyright status of all pre-existing DNA sequences used in their work.

This post will focus on the incorporation of DNA sequence searching into the Copyright Office’s registration function.  Stay tuned for Part 2 of this series, where I plan to discuss the utility of copyright sequence searching for third party users of pre-existing DNA sequences, analogous to the current practice of performing a patent sequence search to assess

Understanding the Differences Between Patent and Copyright Requirements

For readers more familiar with patent prosecution than copyright registration, it might come as a surprise to learn that, as a general matter, the Copyright Office does not perform any sort of novelty search when examining a work that has been submitted for registration.[1]  In contrast, the Patent and Trademark Office (“PTO”) routinely performs sequence searches in order to identify prior art that could undercut the novelty and nonobviousness of a patent claim reciting a DNA sequence.  This distinction reflects the less stringent requirements for copyrightability.

In particular, there is no novelty or nonobviousness requirement for copyright – so long as a work is the product of independent creation and incorporates some minimal degree of creativity, it is copyrightable. No matter how similar (or even identical) it might be to the previous work of another.

Copyright Office’s Response to an Attempt to Register an Engineered DNA Sequence

Interestingly, however, the Copyright Office recently seemed to take the position that a novelty search should be a prerequisite to registration of an engineered DNA sequence.  As mentioned in a previous post, Professor Andrew Torrance of the University of Kansas School of Law and I recently teamed up with DNA2.0, a leading gene synthesis company, in an attempt to register an engineered DNA sequence with the Copyright Office as a copyrighted work.  We appealed the Copyright Office’s decision to deny registration, and received a letter from the Office’s Director of Copyright Policy and Practices affirming the Copyright Office’s position that engineered DNA is not copyrightable subject matter.

The letter provided a number of purported rationales for the Office’s determination that DNA cannot be copyrighted (none of which we found convincing).  Of particular relevance to the current discussion, one of the Office’s justifications was its assertion that it lacked the ability to examine our submitted sequence to assess whether it was truly a work of “creative human authorship,” as opposed to a “naturally-occurring sequence, or a derivation thereof.”  This implies that the Office believes that if engineered DNA was copyrightable, it would be incumbent upon the Office to perform some sort of sequence search to ascertain whether the sequence is truly a work of human engineering, as opposed to an uncopyrightable naturally-occurring sequence, or a derivative of a naturally-occurring sequence lacking sufficient independent creativity to crossover the threshold into copyrightable subject matter.

Search is the Key to Enabling Engineered DNA Copyright Protection

I tend to agree with the Copyright Office on this point, to the extent that I believe that if the Copyright Office begins registering engineered DNA sequences, it should perform a search to confirm the originality of the submitted sequence, i.e., that the sequence is not too close to a naturally-occurring sequence, nor is it too close to another’s previously disclosed engineered sequence.  Even though the standard for copyright is originality, not novelty, in most instances involving a relatively complex sequence, a lack of novelty would (for all practical purposes) establish lack of originality.  As an analogy, while a composer might truthfully assert that she has independently come up with a simple tune that happens to be the same as one already existing in the public domain, an author would not be able to credibly argue that he has independently authored a novel identical to Moby Dick.

However, I strongly disagree with the Copyright Office’s suggestion that its inability to perform novelty searches is a justification for denying registration for engineered sequences.  If engineered DNA is determined to be copyrightable, the Office will be required to provide a mechanism for their registration, just as they are for any copyrightable work.  And if the Office determines that a novelty search should be a prerequisite to registration, it will become incumbent upon the Office to develop the capability for performing such a search.  When inventors began filing patent applications with claims reciting DNA sequences the PTO did not have the internal capability of performing patent sequence searches, but it responded by developing the capability and implementing rules relating to nucleotide and amino acid sequence disclosure, search and examination.[2]

Reforming the Copyright System to Accommodate DNA Sequence Searching

This does not necessarily mean that the Copyright Office would have to develop in-house DNA sequence search capabilities – it could , or perhaps another public agency.  On this point, a group calling itself the Copyright Principles Project (CPP) recently published a White Paper outlining various proposals for reforming the copyright system, one of which is for the Copyright Office to “transition away from being the sole registry for copyrighted works and toward certifying the operation of registries operated by third parties, both public and private.”[3] The CPP describes itself as twenty people having “various kinds of expertise and experience with copyright law and policy, including law professors, lawyers from private practice, and lawyers for copyright industry firms.”  While acknowledging that their proposal to outsource some of the Office’s registry function to third-party contractors might be considered “radical,” they argue that such a move would be justified by the “reality  that the functionality of the [Copyright Office’s] registry remains woefully behind what leading-edge search and database technologies permit.”  The CPP states that “the basic idea…would be to shift the Copyright Office away from day-to-day operation of the copyright registry and toward a role of setting standards for and superintending a system of separate but networked and interoperable private registries.”

Under the CPP proposal, a private company such as GQ Life Sciences could provide the Copyright Office with the DNA search capability it would need to perform novelty searches in conjunction with registration of engineered sequences, and handle the day-to-day operation and maintenance of the searchable database, while the Office oversees the process.  This approach would have the merit of facilitating efficient search and registration of copyrightable engineered sequences, thereby enabling the Office to deny registration of a sequence that appears too close to naturally-occurring sequences, or perhaps to a previously registered sequence.  It would also allow the Copyright Office to take advantage of the well-developed expertise and curated databases maintained by a private search firm such as GQ Life Sciences, rather than requiring the Office to build up its own capabilities in this area from scratch.

[1] Compendium of U.S. Copyright Office Practices (Third Edition, Dec 22, 2014), 602.4(C) No Searches or Comparison of Works (“When examining a claim to copyright, the U.S. Copyright Office generally does not compare deposit copy(ies) to determine whether the work for which registration is sought is substantially similar to another work. Likewise, the Office generally does not conduct searches to determine whether the work has been previously registered.”).

[2]See, Manual of Patent Examining Procedure, Eighth Edition, 2420.   The Requirements for Patent Applications Containing Nucleotide Sequence and/or Amino Acid Sequence Disclosures – the Sequence Rules [R-08.2012]  (describing difficulties in searching and examining nucleotide sequences prior to the implementation of rules relating to nucleotide and amino acid sequence disclosures).

[3] Pamela Samuelson,The Copyright Principles Project: Directions for Reform, 25 Berkeley Tech. L.J. 1175 (2010), Available at: http://scholarship.law.berkeley.edu/facpubs/563.

About Chris Holman

Chris Holman is a Professor at University of Missouri-Kansas City Law School, as well as the author Holman’s Biotech IP Blog and Executive Editor of Biotechnology Law Report. GQ Life Sciences is thrilled to have him be a guest blogger.

Update: GQ-Pat now has over 334 million sequences

Back in July we reported that there were 300 million sequences in GQ-Pat, including 256 million nucleotide sequences and over 45 million protein sequences.  And these protein sequences aren’t just automated translations of nuceotides like TrEMBL. All of these sequences are in fact found in patents and patent applications from patent authorities around the world.

View the latest GQ-Pat Statistics

This year we’ve added 75,000 documents and 27 million sequences, making GQ-Pat even bigger than before.

To put this accomplishment in perspective, when the Human Genome Project formally began in 1990, there were fewer than 40,000 sequences in GenBank, before being transferred from Stanford to the newly created National Center for Biotechnology Information (NCBI).

As of this month, according to the NCBI’s GenBank Statistics page, there are 193 million nucleotide sequences in GenBank/EMBL/DDBJ consortium, the world’s gold standard for sequence databases.*

Updated on a weekly basis, the GenomeQuest database, GQ-Pat, is the most comprehensive and up-to-date IP sequence database available.

Not only does GQ-Pat have more sequences than GenBank, these sequences help searchers in other ways as well:

  1. The sheer size of the database itself helps organizations save money through efficiency and the fact that important search results won’t be missed.
  2. Sequences in GQ-Pat are well annotated because all of them have been found in patents, making them more valuable to researchers. Patent information includes descriptions of a particular invention, including the way in which the invention is used, the inventors, the owners, biological information about the sequence, its function, and so on.
  3. Researchers using GQ-Pat can obtain results much sooner than those using public databases like GenBank because patents are typically filed before publications are drafted.

We’re so pleased that more and more researchers are turning to GQ-Pat to search sequences for a huge variety of life science related projects. When it comes to researching or protecting intellectual property, the quality of the results are often only as good as the size of the database they come from. That’s why we’re dedicated to maintaining the world’s largest. Of course, we’re also adding rich annotations and making sure updates are added on a weekly basis.

If you’ve never tried it, and you’re interested in searching our 334 million sequence (and growing) database for yourself, there’s never been a better time to get in touch about a free trial.

* There are another 50 million protein sequences in Uniprot, the leading protein sequence database, although 98% of Uniprot is TrEMBL, which consists of 49 million unreviewed computer-generated protein translations of nucleotide sequences already in the nucleotide databases.

Try LifeQuest or GenomeQuest Today!

Search Using Natural Language

Search Patents Using Biological Sequences

biosimilar drugsAbout a week ago, the FTC filed a suit against Endo Pharmaceuticals (among others) for blocking generic competition of its Opana ER and Lidoderm products. It is alleging that in addition to paying to delay competitive products, Endo used a no-AG commitment that gave Watson Laboratories more than 7 months of their own monopoly on the market during which Endo would not compete.

Clearly, the opportunities for generic drug products is huge, but there are regulations that must be adhered to when strategizing the best ways to maximize a company’s return on drug development investments or planning an entry into a market with a new biosimilar product.

Take for example, Novartis-owned Sandoz Inc., which released a drug called Zarxio in September of 2015. Zarxio is a copycat version of Amgen’s billion dollar Neupogen drug, and the first FDA approved biosimilar drug launched in the US (the world’s largest pharmaceutical market). It is currently sold 15% below the price of the original drug.

Neupogen and Zarxio are both biological drugs based on the naturally occurring human granulocyte colony stimulating factor. The 175 amino acid protein is used to stimulate the growth of white blood cells in people with compromised immune systems, for example in patients undergoing chemotherapy or bone marrow transplants. Zarxio is a perfect copy of Neupogen with the exact same amino acid sequence, and as can be expected the two drugs are indistinguishable from each other in a wide variety of FDA-administered tests.

Biosimilars have been allowed in the US since 2010 as part of the Affordable Care Act, but it took until May 2015 for the legal dispute between Amgen and Sandoz Inc. to play out. In Europe, which has allowed biosimilar drugs on its markets for almost a decade, eight competing Neupogen copycats have taken almost 80% of the market.

Allowing biosimilars opens up new opportunities in the US pharmaceutical market and understandably, many generic drug companies are trying to get in on the action. Neupogen biosimilar drugs from Apotex and Hospira are already under FDA review, and are expected to enter the US market soon.

Keeping a finger on the pulse of the market for very specific biological sequences is becoming increasingly important. Considering the pace at which products are being developed, it is also becoming more challenging. So what’s an IP researcher to do? For starters, setting up processes and tools to keep better tabs on key sequences in your IP portfolio.

To continue with our Neupogen example, I ran its entire 175 amino acid sequence through GenomeQuest, the largest biological sequence IP database*. Here is the Neupogen sequence in case you’re interested in replicating my search or perhaps want to start your own generic pharma company.

sequence search

The search results revealed that there are 366 applications that have this exact same sequence, falling into 101 extended INPADOC families, owned by about as many companies. A quick analysis of the results also tells us that a staggering 31 of these documents are granted and mention the sequence in the claims. There is no doubt in my mind that everyone who has a horse in this race is well aware of these documents and has looked at them carefully.

So why go over this example? As the volume of research required to compete increases, the efficiency and thoroughness of getting a complete set of results will be a significant competitive advantage. Using GenomeQuest, the sequence search and analysis tasks took only about five minutes, and I was very confident that I had found all the important documents. The only additional step I would recommend would be a sophisticated full-text keyword search that could bring up more, but I doubt anyone would write an application relevant to this biological without including the sequence itself.

Want more inspiration? An even bigger bounty lies in copying multibillion dollar antibody drugs for autoimmune diseases and cancer. In February 2015 South Korea’s Celltrion launched a biosimilar version of Johnson & Johnson’s monoclonal antibody, Remicade, in Europe. Remicade is an antibody that is used to treat autoimmune diseases like rheumatoid arthritis and inflammatory bowel disease. The FDA is currently reviewing Celltrion’s biosimilar antibody for release on the US market.

Searching for antibody biosimilars is a little more involved. What matters most are the Complementarity Determining Regions (CDRs), small regions in the antibody amino acid sequence that interact directly with the specific antigen target. It is possible to change a lot of the amino acids outside of these CDRs to possibly avoid infringement, and still have a working biosimilar drug. The good news is that GenomeQuest not only provides access to all the data, it also incorporates the search algorithms and analysis capabilities that help you take those variations in amino acids outside the CDRs into consideration so you can generate better results faster.

*GenomeQuest contains over 315 million sequences found in 710 thousand patent applications from all over the world. Learn more about GenomeQuest.

Try LifeQuest or GenomeQuest Today!

Search Using Natural Language

Search Patents Using Biological Sequences

It’s hard not to notice that the future is arriving faster than ever. Innovations are arriving at breakneck speed. In just the last couple of years we have seen things like self-driving cars, software that beats Jeopardy!, bipedal robots that can navigate difficult terrain just like humans do, and a very convincing defeat of one of the world’s best Go players by Google’s DeepMind technology. These are all things that knowledgeable people predicted would still be decades away, yet they are all happening right now. Why is that, and what does that have to do with your IP strategy?

Moore's LawMost of us would agree that digital technology is a big, if not the biggest contributor to the accelerating pace of innovation. By now, everyone should be familiar with Moore’s law, what it has done to processing power, and how this has impacted every aspect of our lives. It’s been more than 50 years since Moore first published his “law,” and it is still going strong. However, what a lot of people may not realize is that we’re only now getting to the incredibly fast growth part of the exponential curve. This means that the speed of progress that we observed in the past will no longer be a reliable indication for what will happen in the future. In other words, the acceleration itself is accelerating, and it’s happening so fast, that it’s forcing organizations to adjust the way they do business.

Nowadays, Moore’s law applies to more than just computational power. Other technologies have been miniaturized and transferred to silicon, thereby becoming subject to a similar pace of digital progress. These technologies include microfluidics and the wide variety of sensors that are driving life science technologies like increasingly efficient whole genome sequencing, high-throughput screening of drug compounds, and combinatorial chemistry.

Of course, hardware isn’t the only technology that is responsible for these changes. There is a mountain of freely available data that is growing at a staggering speed. It is mind-blowing to think that over 90% of all digital data in existence today has been produced only in the last two years. Ranging from the latest Neanderthal genome to the holiday pics that my Uncle Joe posted on his Facebook page this morning, the volumes of data that are available are opening doors to more measurement-driven decision making. For example, the frequency of specific google search terms is now a much more reliable indicator for upcoming house price changes in an area than the expert opinions of local real estate agents.

We’ve all witnessed the advancements in technology, and we’re all trying to get a grip on the incredible increase of readily-available data. But how exactly does this lead to more and faster-paced innovation? The quick and simple answer is that it lowers the costs of entry dramatically, making it is easier to do very targeted experiments. However, there’s more to it than that.

Innovation often takes place at the intersections of existing technologies. When the people at Waze combined a digital map with GPS navigation and sensor feedback from millions of mobile phones, they got real-time information about what is happening on the road. This allowed them to suggest the best way to travel for you at the very moment you need it, taking into account current traffic flow and road accidents. (In 2013 Google paid almost a billion dollars for this idea). Sure, Waze didn’t invent rockets, satellites, GPS, digital maps, or smartphones, but they did create a clever combination of already existing technologies.

Likewise, the people at IBM are now feeding Watson (the machine that won Jeopardy!), medical textbooks and literature so that it can assist with difficult medical diagnoses. One can only imagine what will be possible when software like this is able to run on ordinary computers and becomes available to the public. Would you feed it MedLine and the world’s patent literature just to do your next prior art search?

In addition to more combinatorial innovations based on past accomplishments, another promising effect is that we may also have more people to do the actual innovating. In 2011, when Sebastian Thrun from Stanford University decided to put his introduction to artificial intelligence online, over 160,000 people from 190 countries signed up. While not all of them finished the course, out of the ones who did, 400 of them had better grades than the best Harvard student doing the same course in person. High quality online education now has the brightest minds of the entire planet working on interesting problems, not just those that are lucky enough to have access to traditional education.

So what does that mean for your IP strategy? Of course, patents will continue to protect and commercialize ideas. However, new ideas and opportunities are going to continue to arrive at a much faster pace, and innovators will have to become more efficient at recognizing and managing them, or risk missing out.

The potential cost of missing opportunities is probably what should worry organizations the most. There is less and less room for second and third best place in today’s world. New technologies allow delivery of goods and services on a worldwide scale, but they also allow people to be very selective. Online reviews and comparisons help them identify the best solutions, services, and products instantly. To take an example from the consumer space, where once a restaurant could get away with lousy food because a nearby tourist attraction supplied an endless stream of first-time customers, Yelp now supplies data from the public that forces a restaurant owner to become better or get forced out of the market.

Although the acceleration of innovation is leading to an ever-increasing number of product and services, in a growing number of markets, the momentum created by various forms of data (including user data) is creating more and more winner-takes-all scenarios. There’s only room for so many good navigation apps, and even when the 10th most popular app does most of the things that Waze does, it will not be worth anywhere near a billion dollars.

Why would you expect customers to settle for anything less than the very best solution, when all relevant information is easy to find online? When this happens in your industry, (and it is almost inevitable that it will), you want to be the one with the best solution, or at least be holding a part of the relevant IP.

The better your organization is at visualizing and processing the bounty of opportunities that can be found in the sea of data and exciting technologies heading our way, the better prepared you’ll be for what will come next. At GQ Life Sciences, we’re constantly improving and adding to our search products and databases to help our customers identify relevant IP and prior art more efficiently and more accurately. Want to get ahead of the curve and start monitoring your current portfolio? Just click the button below to sign up for a free trial.

Try LifeQuest or GenomeQuest Today!

Search Using Natural Language

Search Patents Using  Biological Sequences

Lawyer searching patentsWith the United States Supreme Court ruling in Myriad, the enforceability of certain claims in existing gene patents and the broader patentability of genetic material has been cast into a state of some disarray.

In this three part series, guest blogger and thought leader Professor Chris Holman at the University of Missouri-Kansas City Law School explores the idea that copyright law rather than patent law may better suit genetic material as intellectual property. Professor Holman authors the well-known Holman’s Biotech IP Blog and is the executive editor of Biotechnology Law Report.

Part 3. [Please see Part 1 and Part 2 should you have missed it.]

Since publishing my DNA copyright article in 2011, I have been approached by a number of companies who are interested in exploring the possibility of using copyright to protect their products.  I have also been contacted by a number of attorneys at intellectual property law firms who are considering the applicability of copyright to engineered DNA.  In fact, I am aware of at least one law firm that is actively making presentations to major biotechnology companies encouraging them to consider asserting copyright in their products, presumably by filing a copyright infringement lawsuit against an alleged infringer.  Recall that registration with the Copyright Office can strengthen copyright, but it is not a prerequisite to filing and winning a copyright infringement lawsuit (as long as the copyright owner at least attempts to register the work).  So the Copyright Office’s position on the copyrightability of DNA does not preclude the courts from declaring engineered DNA copyrightable.  There is precedent for this.  For example, the Copyright Office originally refused to register videogame displays, asserting that they were not copyrightable works, but this was appealed and the courts overruled the Office, finding that in fact a videogame display could be copyrightable.

Indeed, one biotechnology company is sufficiently interested that it agreed to collaborate with Professor Torrance and myself in a project designed to advance the conversation on the copyrightability of DNA.  The company is DNA2.0, a leading DNA synthesis company that designs and synthesizes engineered genetic sequences.  Professor Torrance and I assisted DNA2.0 in attempting to register one of the company’s engineered DNA sequences with the Copyright Office as a copyrightable work.  As expected, the Office refused to register it, and we responded by filing petition arguing for the copyrightability of engineered DNA.  Our petition pointed out some of the policy advantages which could flow from such a development.  We were not optimistic on the likelihood of changing the Office’s position, but at least we hoped to get a formal statement of the Copyright Office’s rationale its position on this issue.

It took 14 months to hear back from the Copyright Office, but when we did we were gratified to see that we had received a detailed explanation from the Office’s Director of Copyright Policy and Practice, who apologized for the delay, explaining that this was a matter of first impression for the Office and that Office personnel had spent a substantial amount of time figuring out how to respond to it.  Frankly, in our opinion none of the Director’s legal justifications for the disparate treatment of engineered computer code and engineered genetic code were persuasive.

We could have appealed the case in the courts, as was the case with videogame displays, but the cost of pursuing such an appeal was more than DNA2.0, still a relatively small startup, was willing to commit.  I have written up an article describing our experience attempting to register the DNA2.0 sequence, including a detailed response which we think refutes the Office’s stated rationale for denying registration, which should be published sometime in 2016.  I think many in the biotechnology community will find it interesting.  Who knows, maybe at some point there will be enough interest that someone will actually challenge the Copyright Office’s stance on engineered DNA in the courts.  It would be very interesting to see how a court would respond – frankly, I think the Copyright Office would have a hard time defending its position, much as it did when it argued that videogame displays cannot be copyrighted.

About Chris Holman

Chris Holman is a Professor at University of Missouri-Kansas City Law School, as well as the author Holman’s Biotech IP Blog and Executive Editor of Biotechnology Law Report. GQ Life Sciences is thrilled to have him be a guest blogger.

Try LifeQuest or GenomeQuest Today!

Search Using Natural Language

Search Patents Using  Biological Sequences

shutterstock_306810023With the United States Supreme Court ruling in Myriad, the enforceability of certain claims in existing gene patents and the broader patentability of genetic material has been cast into a state of some disarray.

In this three part series, guest blogger and thought leader Professor Chris Holman at the University of Missouri-Kansas City Law School explores the idea that copyright law rather than patent law may better suit genetic material as intellectual property. Professor Holman authors the well-known Holman’s Biotech IP Blog and is the executive editor of Biotechnology Law Report.

Part 2. [Please see Part 1 should you have missed it.]

… Given the similarity of computer code and genetic code, why has copyright been extended to one but not the other?

In part, I think the answer lies in the fact that in the 1970s, as the development of computer software was first coming to be recognized as an economically significant enterprise, it was far from clear whether patent protection would be available for computer programs. Indeed, two Supreme Court cases decided in 1972 and 1978, Gottschalk v. Benson and Parker v. Flook, held that the computer programs at issue in those cases were not patent eligible. Given the substantial investment required to develop commercially relevant computer programs, and the ease with which the resulting product could be copied and illicitly disseminated, some form of intellectual property was deemed necessary. Patents are the form of IP traditionally associated with technology, but if patents were not going to be available for computer programs, then perhaps copyright could be enlisted. A number of copyright experts, particularly legal scholars, argued vehemently at the time that copyright should not be extended to software, but ultimately the Copyright Office began registering computer programs under a “rule of doubt” in the 1960s, and in the 1970s and early 80s the courts issued a number of decisions upholding the copyrightability of computer code, even highly utilitarian software written in a form that can only be read by computers. In 1981 the Supreme Court decided Diamond v. Diehr, upholding the patentability of the computer program at issue in that case, and in the 1980s and 1990s relatively expansive patent protection became available for computer programs, which led to the current coexistence of patent and copyright protection for software.

In contrast, since the earliest days of biotechnology patents have been viewed as an effective means of protecting biotechnological innovation, a belief fostered by the Supreme Court’s 1980 decision in Diamond v. Chakrabarty which held that genetically modified living organisms can be patented, and generally espousing an expansive view of the scope of patentable subject matter. With the availability of patents, there was no compelling need to push for an alternative form of protection such as copyright. Furthermore, the earliest products of biotechnology tended to involve relatively simple genetic constructs comprising only slightly modified versions of naturally-derived sequences, thus looking less like an engineered sets of instructions than does computer code.

I can imagine at least two types of bioengineered sequences with respect to which copyright might be applicable. One would be synthetic protein-coding sequences, i.e. synthetic genes. Each codon of such a sequence represents an instruction directed towards a biological machine – start or stop translation, or incorporate a specified amino acid into a growing peptide chain. As an ordered set of instructions directed towards a machine, it should be copyrightable for the same reasons a computer program is, assuming it meets the various requirements for copyrightability such as originality, creativity, and fixation, as discussed below and in greater detail in my DNA copyright law review article.

The second category of potentially copyrightable genetic works would include constructs made by combining multiple genetic coding elements, as exemplified by the Smolke lab’s engineered opioid synthesis construct. Professor Endy has written extensively on the increasingly modular structure of the products of synthetic biology, which he points out is analogous to the modular nature of computer programs, as well as other more traditional works of engineering. Even a construct comprising nothing more than a unique reconfiguration of naturally occurring genetic elements could be afforded copyright protection, at least if that configuration exceeds some threshold level of originality and creativity. It is well-established that an original “selection, arrangement, or coordination” of uncopyrightable elements can result in a copyrighted work.

In the context of software, copyright is particularly effective as a tool for going after pirates, i.e., those who takes advantage of the ease with which computer code can be copied and seek to profit from the commercialization of unauthorized copies. In the same way, I think that copyright could play an important role in policing against the pirating of bioengineered products. Genetically modified seeds are one type of product that I can imagine really benefiting from copyright protection. Like software, genetically modified seeds are highly vulnerable to piracy, given the ease with which the product can be used as the template for the production of a virtually unlimited number of unauthorized copies. Patent owners like Monsanto currently rely heavily on patents to go after farmers who illicitly save patented seeds to plant the next season, or to sell to others. There could be a number of practical advantages to the use of copyright in conjunction with, or as an alternative to, patents, much as is the case today with respect to software. The enforcement of copyright can be more straightforward – proving copying in the case of piracy, for example, should be easier in many cases than proving patent infringement, and dealing with a defendant’s claims of invalidity or unenforceability would be less of a burden for a copyright plaintiff than is the case with patents. The remedies available for a prevailing copyright owner can also be attractive, and there is even the possibility of pursuing criminal enforcement of copyright in particularly egregious instances of piracy, or of blocking importation.

In fact, there are a number of advantages of copyright over patent that could be of benefit to biotechnology innovators. Obtaining patent protection is notoriously expensive and time-consuming, and often takes years after the initial filing of the patent application. Copyright, in contrast, exists from the moment a copyrightable work is “fixed in a tangible medium of expression.” In other words, a copyrightable DNA sequence would be protected by copyright from the moment it is first embodied in a DNA molecule, or recorded as a string of letters on a piece of paper or on a computer-readable medium like a hard drive. Unlike a patent, there is no need to seek government approval for a copyright. It is possible to register a copyrighted work with the Copyright Office, and there are a number of advantages to doing so, but the copyright exists at the time of fixation regardless of whether the work is registered or not. Thus copyright could be an attractive alternative for a company that is generating a large number of synthetic DNA sequences for which patent protection would be prohibitively expensive. The availability of protection from the time of fixation would also be an advantage with respect to a company seeking to quickly commercialize a product based on engineered DNA, perhaps months or even years before a patent would issue.

In many ways copyright protection would be substantially thinner and less robust than patent protection, and indeed that is one of its main advantages, since it could provide some meaningful protection without unduly hindering subsequent research and innovation by others. There are those that complain that biotechnology patents, especially so-called “gene patents,” threaten to create a thicket of property rights that impede basic research and follow-on innovation, and this seems to have been a concern that the Supreme Court sought to address in Myriad. In fact, there are many critics of software patents who argue that the more limited rights of copyright are a preferable form of intellectual property for computer programs, and for much the same reasons the same could be the case for some, although by no means all, engineered DNA products.

For one thing, copyright infringement only occurs if one actually copies a copyrighted work. Independent creation is an absolute defense, which is not the case with patents.

It is also significant that copyright would not cover the functionality of engineered DNA, only the specific copyrighted DNA sequence, and perhaps some relatively similar sequences. This is a direct consequence of one of the most foundational principles of copyright, which lawyers refer to as the “idea-expression dichotomy,” –i.e., copyright does not extend to ideas, but instead is limited to a specific means of expressing the idea. As applied to engineered DNA, the idea-expression dichotomy essentially prevents copyright from tying up the functionality (the “idea”) of an engineered DNA sequence, but can only be used to preclude others from copying the specific sequence used by the copyright owner to achieve that functionality (the “expression”), and probably some relatively narrow set of highly similar sequences. Thus, any third-party would be free to copy the functionality of an engineered DNA sequence, so long as that third-party is willing to invest the effort in coming up with an alternative DNA sequence for achieving that function. In the case of a synthetic gene, for example, the redundancy of the genetic code should allow for a virtually unlimited number of non-infringing alternatives. Importantly, the relative thinness of copyright does not diminish its utility as a means for going after pirates, as illustrated by the success with which suffer companies have been able to wield their copyright.

There are a number of limitations on the scope of copyright protection afforded software that flow directly from the idea-expression dichotomy, and which by analogy would be applicable to engineered DNA. For example, under what is referred to as the “merger” doctrine, if there are only a limited number of ways to achieve a particular function, those expressions cannot be copyrighted, out of concern that any copyright might threaten to impermissibly restrict access to the functionality. Furthermore, elements of a computer program that are derived from the public domain, or dictated by considerations of efficiency, or dictated by external factors (such as requirements for interoperability with hardware or with other programs) are also unprotectable, and freely available to subsequent innovators. As applied to engineered DNA, these limitations on copyright would render it much less likely that copyright would limit subsequent innovation than the current patent-centric approach.

Copyright law also incorporates a variety of defenses, exemptions, and compulsory licensing provisions that could provide more breathing room for subsequent research and innovation. For example, the copyright statute provides a defense of “fair use” which can shield those who engage in socially beneficial or non-commercial forms of infringement from liability. In the context of engineered DNA, fair use could be enlisted to protect academics and other non-commercial actors from liability. It would permit even commercial actors to make copies of a copyrighted sequence for legitimate purposes, such as to understand the underlying functionality of a sequence. The copyright statute also provides a number of subject matter-specific exemptions from liability, as well as compulsory licenses, including 17 USC 117, which limits the enforceability of copyright on computer programs. If copyright on DNA were found to have some unintended negative consequences, there is ample precedent for Congress to address them by DNA-specific amendment of the copyright statute.

For years Professor Endy and others have advocated for “open source biology,” hoping to build on the success of the open source software movement. But as of yet, open source has never really taken off in biotechnology the way it did with computer programs, and I think one of the obstacles to open source in biotechnology is the current patent-centric approach to protecting biotechnological inventions. The intellectual property that forms the basis for open source software is copyright, not patents, and I think that if synthetic biology could be incorporated into the copyright system it would facilitate the growth of true open source synthetic biology. One of the fundamental problems with trying to use patents as the basis for open source is that licensing a patent does not confer freedom to operate, since the license does not rule out the possibility of infringement of a third parties patent. In contrast, since a copyright is only infringed by copying, a follow-on innovator could incorporate a copyrighted sequence into her own work without fear of infringing the copyright of another, so long as the incorporation was authorized by the owner of the copyright in the incorporated sequence.

In spite of the compelling logic supporting the recognition of copyright for engineered genetic code, and the potential policy benefits of providing a viable alternative to patents for synthetic biology, it seems unlikely that copyright will be extended to DNA anytime soon. The Copyright Office has officially taken the position that DNA is not copyrightable, and to my knowledge the issue has never been presented to the courts. It is certainly not on Congress’s radar. But I sense an increasing interest among biotechnologists and intellectual property attorneys, and I suspect that at some point the issue will be before the courts.

[keep an eye out soon for the third blog in this series…]

About Chris Holman

Chris Holman is a Professor at University of Missouri-Kansas City Law School, as well as the author Holman’s Biotech IP Blog and Executive Editor of Biotechnology Law Report. GQ Life Sciences is thrilled to have him be a guest blogger.

Try LifeQuest or GenomeQuest Today!

Search Using Natural Language

Search Patents Using  Biological Sequences

Copyright ImageWith the United States Supreme Court ruling in Myriad, the enforceability of certain claims in existing gene patents and the broader patentability of genetic material has been cast into a state of some disarray.

In this three part series, guest blogger and thought leader Professor Chris Holman at the University of Missouri-Kansas City Law School explores the idea that copyright law rather than patent law may better suit genetic material as intellectual property. Professor Holman authors the well-known Holman’s Biotech IP Blog and is the executive editor of Biotechnology Law Report.

Drew Endy, a professor at engineering at Stanford University, has been called “synthetic biology’s most compelling evangelist” (Michael Specter, Denialism: How Irrational Thinking Hinders Scientific Progress, Harms the Planet, and Threatens Our Lives) and is known for innovative ventures such as the Biobricks Foundation and Registry of Standard Biological Parts.  During a 2012 keynote address at a Stanford Law School Conference on Intellectual Property Law and the Biosciences, Professor Endy opined that “given the history in software, there is going to be for the foreseeable future an ever-renewing enthusiasm for exploring the idea of copyright” for synthetic biology.  He noted that “literally every student I see ….  who connects with property rights immediately presumes that you should be treating this stuff like code, and they are familiar with using copyright in that context.”  Indeed, given the broad availability of copyright protection for software, and the growing analogy between engineered genetic code and computer code, it has long struck me as anomalous that copyright for engineered DNA has yet to be given more serious consideration.  The potential applicability of copyright to engineered DNA has long been noted, particularly by a small number of academics who understand the theoretical underpinnings of copyright protection for software and recognize the similarity between genetic and computer code.

The first formal treatment of the subject of which I am aware is Copyright in Living Genetically Engineered Works, a law review article by Professor Irving Kayton published in 1982.  Professor Kayton is well known to patent attorneys as the founder of the Patent Resources Group (PRG), the original provider of patent bar review courses – I met Professor Kayton in the mid-1990s while I was taking the PRG patent bar review course in Los Angeles, my first real exposure to patent law while I was still working as a postdoc.  But in the early 1980s he was giving a CLE presentation on copyright law, and one of the attendees asked him whether, given the copyrightability of computer programs, engineered DNA sequences might be considered copyrightable subject matter.  Kayton admits that he was initially “shocked and perplexed” by the suggestion, but as a good scholar he could not simply dismiss the idea out of hand, and the more he thought about it the more convinced he became that in fact, if software is copyrightable, there is no valid legal basis for excluding engineered DNA from copyright protection.  He wrote up the results of his legal analysis in the prescient 1982 article, which concluded that “[u]nder certain circumstances, from a practical as well as legal viewpoint, copyright protection may be the only or the most effective way an ‘author’ can protect a valuable genetic ‘work.’”

Since then a number of other scholars have come to a similar conclusion – two of the most thorough treatments of the subject of which I am aware were independently published in 2011 by myself and Professor Andrew Torrance of the University of Kansas (a law professor who, like myself, practiced as a biotechnology patent attorney prior to entering academia).  See Christopher M. Holman, Copyright for Engineered DNA:  An Idea Whose Time Has Come, West Virginia Law Review, Vol. 113, pp. 699-738 (2011); Andrew W. Torrance, DNA Copyright, Valparaiso Law Review, Vol. 46(1), pp. 1-41 (2011).

I find that most people are initially skeptical of the notion that DNA could be copyrighted, often pointing out that DNA is functional, and assuming that copyright is reserved for more aesthetic works such as music, art, and literature.  But the fact is that although copyright has historically been associated with non-utilitarian, aesthetic works, any implicit prohibition against copyright for functional works was shattered in the 1970s and 80s when copyright protection was extended to wholly utilitarian computer programs that are only interpretable by a machine, e.g., a string of zeros and ones encoding instructions for how to manage the air-fuel ratio in an automobile.  In other word, the great conceptual leap occurred more than 30 years ago when the law began treating functional computer code as a copyrightable “literary works” – by comparison, a further extension to encompass engineered genetic code would be relatively minor and incremental.

It is important to emphasize that I am not talking about copyright for naturally-occurring sequences, or even sequences that vary only slightly from a naturally-occurring precursor.  Copyright would not create property rights that would encompass people’s genes, for example, a concern I often hear raised, analogous to the pervasive myth that patents somehow allow biotechnology companies to “own” people’s bodies.  As explained in my West Virginia Law Review article, copyright law requires originality and some minimal level of creativity, which should effectively bar copyright protection for genetic sequences that come too close to anything that occurs naturally.  In fact, it is conceivable that copyright law could require a more substantial departure from nature than patent law, depending upon how the courts interpret the requirements of creativity and originality in the context of DNA.  Bear in mind that prior to the Supreme Court’s recent Myriad decision patents were routinely granted on DNA sequences identical to naturally-occurring genes, and even post-Myriad all that appears to be necessary for patent eligibility is a “marked difference” from nature (whatever that means).

In order to fully appreciate my argument for extending copyright to engineered DNA, it is necessary to understand the underlying basis for permitting copyright on computer programs, and then recognize the functional similarity of computer code and genetic code.  A computer program is essentially a set of written instructions directed towards a machine that direct the machine to perform a series of functions, and the law has come to accept this set of instructions as a copyrightable “literary work.”  An engineered DNA sequence is likewise a set of written instructions directed towards a machine, albeit a biological machine, which directs that machine to perform a series of biological functions.  A typical gene sequence, for example, is a set of instructions directing a cell to incorporate amino acids in in a specified order.  At a larger scale, an engineered genetic construct can encode a complex set of instructions involving the regulation of expression, feedback loops, and the like to achieve increasingly sophisticated functions – the sort of complex bioengineering commonly referred to as synthetic biology.  A nice example is the recently reported engineering of a 23 gene opioid production pathway in yeast by Christina Smolke’s group at Stanford.  See Galanie, S. et al., Complete biosynthesis of opioids in yeast, Science 349, 1095–1100 (2015).

Interestingly, it is not uncommon for today’s synthetic biologists to refer to genetically modified cells as “machines” and genetically engineered DNA as “code,” and to the use of this code to “program” cells.  As I explained in another law review article a few years back, the engineering of computer code and genetic code are increasingly converging as biotechnology transitions towards synthetic biology.  See Christopher M. Holman, Developments in Synthetic Biology Are Altering the IP Imperatives of Biotechnology, Vanderbilt 17 J. Ent. & Tech. L. 385 (2015).  It is not surprising that Professor Endy’s students, having experience in both software engineering and synthetic biology, are struck by the disparate treatment under copyright law.  Given the similarity of computer code and genetic code, why has copyright been extended to one but not the other?

[keep an eye out soon for the second blog in this series…]

About Chris Holman

Chris Holman is a Professor at University of Missouri-Kansas City Law School, as well as the author Holman’s Biotech IP Blog and Executive Editor of Biotechnology Law Report. GQ Life Sciences is thrilled to have him be a guest blogger.

Try LifeQuest or GenomeQuest Today!

Search Using Natural Language

Search Patents Using  Biological Sequences

Australian DNA ImageAfter the United States Supreme Court ruling in the Association for Molecular Pathology v. Myriad Genetics in June of 2013, the industry scurried. The Court ruled that naturally occurring DNA is not patent eligible even if isolated, but cDNA or “complementary DNA” is because it is not naturally occurring but rather a product of the laboratory scientist even though it is exactly the same nucleic acid information.

In fact, however, cDNA is naturally occurring, as many viruses convert their viral RNA into mRNA through cDNA: viral RNA → cDNA → mRNA. Unfortunately, the US Supreme Court ruling, based in a lack of understanding of the science, created the broadest possible interpretation of isolated DNA and cDNA. This ruling forced the USPTO to create even more complex guidelines to determine whether inventions were patent eligible, which opened the door for dozens of law suits on the eligibility of claims formerly granted pre-Myriad.

Australia recently undertook the same challenge – in this case the Australian High Court took on D’Arcy v Myriad Genetics and ruled similarly that Myriad’s claims relating to isolated BRCA1 nucleic acid were not patent eligible subject matter. However, in Australia’s case, the clear, scientifically grounded, well-reasoned ruling enabled the Australian Patent Office to release narrower, new guidelines that promote innovation by setting clear rules.

In particular, they note that any naturally occurring subject matter which merely claims genetic information, be it isolated DNA, RNA, whether human or non-human, coding or non-coding, is excluded as patentable. In addition, they point out that any man-made constructs that do nothing more than replicate genetic information of a naturally occurring organism are also excluded, including cDNAs, probes, etc. They clarify that this excludes man-made constructs that are not naturally occurring, be they chimeric DNA or novel antibody sequences. This alone makes it much cleaner than the current US system.

In addition to clear specifications of what is excluded, the High Court took things even further by also clarifying what can be considered as patent eligible. These include:

  • Recombinant or isolated proteins.
  • Pharmaceuticals and other chemical substances.
  • Methods of treatment.
  • Methods of applying herbicides.
  • Applications of computer technology.

As we followed the developments in Australia, we thought it might be interesting to  examine trends in patenting to see if there are any recent effects. We generated a heat map using a combination of LifeQuest and GenomeQuest data to better visualize what is happening. It depicts all Australian grants and applications that contain any one of the phrases cDNAisolated DNAisolated protein, or recombinant protein occurring in the claims.

Our findings indicate that the High Court ruling affects a significant number of documents that refer to cDNAs during the ten year span from 1998 to 2008, and that the claimed subject matter for recombinant proteins has continued to grow in recent years. There is a significant, clearly identifiable increase in 2015 of the term cDNA in claims made up of a large number of new applications with rewritten claims that still refer to cDNAs but it is important to note that they are directed toward methods for using them in diagnostic and therapeutic contexts.

Australian blog heat map

The message is clear: with clear guidelines for subject matter provided by the Australian High Court and Patent Office, innovators were quick to see the writing on the wall and reformulate their IP in Australia.

Try LifeQuest or GenomeQuest Today!

Search Using Natural Language

Search Patents Using  Biological Sequences

Infographic: 7 Reasons WhyOur post on the top reasons why scientists should be searching patents was so popular, we’ve put together an infographic of the top seven reasons why patents simply can’t be ignored. It’s a whole new year, so if you or someone on your team is still on the fence about whether to include or expand patents’ role in your research efforts, we hope you’ll take these points into consideration.

For example, did you know that you might be missing 80% of the current technical knowledge if you’re not including patents when examining FTO or monitoring the competitive landscape? [that’s #2 on the list!]

LifeQuest makes it simple to incorporate full-text, natural language patent searches, delivering and organizing your results for more efficient sharing and analysis. As a matter of fact, 84% of the life science patents in LifeQuest extend upon the knowledge found in scientific articles in PubMed. [#4 on the list!]

To see the remaining reasons, download the Infographic. Send it to your boss – send it to your patent research team – or download it just to check out the cool graphics.

Try LifeQuest or GenomeQuest Today!

Search Using Natural Language

Search Patents Using  Biological Sequences

Three Pro Tips for Scientists When Researching PatentsThose of you who have read my previous blog post 10 Reasons Why Research Scientists Should Patent Search know that I use patents to find information about my research topics. One of the most valuable resources I had when I first started researching patents was a lengthy (and rather technical) introduction to the way they work.

At GQ Life Sciences, we all believe that everyone should have easy access to patent information. In that spirit, I’ve summarized three of the most important elements to help you access patent information as efficiently as possible.

  1. Know about the patent life cycle.

You have to understand that most applications for patents are never granted! When you search a patent database, you’ll be searching both granted patents and the large number of patent applications that are in process.

A patent application is born when an inventor submits it to a patent office. A patent examiner, employed by the patent office, decides whether the invention can be patented by comparing it to everything that is already known, the “prior art,” and by determining that the patent is “non-obvious” and has some useful application.  Only after this step is successfully completed does the application turn into a real granted patent that can be enforced. This determination can take years, with multiple back-and-forths between the inventor and the patent office, and the application itself changes through that process, with specific embodiments of the invention being added or removed based on feedback from the patent examiner. During this time, patent databases will describe the patent’s legal status as an “application.”

Once granted by the patent office – where its legal status will become “grant” – it has a roughly 20 year lifespan of protection, after which the legal status becomes “lapsed.” The 20 year term period starts at the priority date, the date the invention was made. Knowing this basic lifecycle can help you determine a lot about the invention being described and its potential impact on your own patent research.

Pro tip: find the most relevant patents by searching only for those documents where the legal status is “grant” or where it is “application” and the publication date is within the past five years.

  1. Understand patent families.

Most applications and patents come in families, a group of patents that are related to each other by a priority date. Take for example the same application filed in multiple countries through WIPO (World Intellectual Property Organization). WIPO doesn’t grant patents, but they will give an opinion on how potentially patentable an invention is. They also give you a priority date, so that if you later decide to file the patent application in one of their 145 member states you get to use that date as the date the invention was made. Taking this route avoids the costs and troubles of filing in lots of different countries/languages right away and still gives you lots of options to protect your invention if it turns out to be valuable later on. Because these WIPO applications all share the same priority date, they end up in a patent family together.

Another way in which US applications can end up in a family together is if they are somehow derived from an earlier application and claim the priority date of that earlier application. This can happen when new claims are added for the same invention (continuation), when the invention has been improved by the same inventor (continuation-in-part), or when an existing application is split into smaller parts (divisional). See more complete explanations of these concepts.  When claims from different applications are mixed into a new application like this, the document can end up with multiple priority dates.

There are two main types of patent families. The first one, often called a “main,” “simple,” “tight,” or “basic” family, is where all documents in the family have the exact same set of priority dates. These are typically the exact same application in different countries. The second one, often called the “extended” family, is where every document shares at least one priority date with at least one of the other documents in the family. These extended families include all the US derivative patents explained above, and can grow much bigger.

Pro tip: use patent families to avoid reading the same document issued in different patent authorities. Instead, pick the single representative document of that family that matters most to your patent research, based on its date, its country, its completeness, or its native language. Also, use patent families to keep track of how a specific invention progresses after the first application has been filed.

  1. Pick your search strategy well.

With an estimated 80 to 100 million patents out there, it is essential that you pick the right strategy to go through them.

You have to find the right keywords to search with first. Unlike scientific articles, inventors often willingly make their patents hard to find by avoiding the usual keywords associated with their domain! They often resort to vaguely describing things instead of outright naming them. In addition, the vocabulary associated with a technology often changes as a research field progresses over time. To find these things you have to cast a wide net first by trying different keyword combinations, using synonym lists, and scanning some of the returned documents to find more and better keywords.

Once you’re satisfied with the search results, it’s time to narrow them down by removing all documents that you aren’t interested in. If you have a lot of results, you can try filtering on specific fields like title or abstract. If one of these contain your keywords the documents are almost guaranteed to be interesting to you. You can also filter by publication date to see only the most recent documents. Most professional searchers I know will even go through a long list of documents one at a time and see why they picked it up. Good software for result analysis with a user interface designed for this task will help tremendously.

Be mindful where you search. In this day and age it’s relatively easy to find patent search services online. Just know that not all of them are easy to use, and that none of them are very complete. For example, the USPTO site is notoriously user unfriendly and their full text archive is only US and doesn’t go further back than 1976. Google patents doesn’t allow you to search in specific fields like title, abstract, claims, or dates. It also has holes in its data. Especially recent documents are not always available immediately.

Pro tip: the one field where inventors have to be specific about their invention is the “Claims” field. It’s the place where they specifically itemize the various components of the invention that they are trying to protect. Make sure you focus key elements of your search on that field.

Try LifeQuest or GenomeQuest Today!

Search Using Natural Language

Search Patents Using  Biological Sequences