March 6, 2014

Will the Genome Compiler and DNA2.0 Settlement Bring Relief to Synthetic Biology Industry?

Maxx Chatsko, SynBioBeta

The synthetic biology industry is relatively new, growing extremely quickly, and pioneering innovative solutions to difficult problems at an unprecedented rate. That, of course, is the rosy view. Every high-growth industry faces challenges; some of which are unique, such as scaling biology, and others that are less discriminatory, such as patent litigation. Not all patent disputes are created equally, however. A disputed production process for a particular chemical will likely only affect a limited number of companies, but a dispute over broader technology would cast a wider net over the industry.

That’s what happened between gene synthesis company DNA2.0 and bioCAD developer Genome Compiler, which enlisted lawyers to defend and challenge, respectively, intellectual property covering features of gene design and assembly software. At worst, the case could have put the entire crop of emerging DNA design software platforms from Genome Compiler, TeselaGen, Genabler, Benchling, and others on notice — possibly bottlenecking virtual design capabilities for the entire synthetic biology industry. At best, the fears of software startups would blow over without interrupting investments or user growth.

Good news: the two sides have reached a settlement a little over one year after the original suit was filed. What does the quick and amicable end to a closely-watched dispute mean for the future of bioCAD software platforms and the industry at large?

How did this get started?

In December 2012 DNA2.0 filed suit against Genome Compiler for allegedly infringing on US Patent 7,805,252 titled “Systems and methods for designing and ordering polynucleotides”. DNA2.0 CEO Jeremy Minshull described the importance of the patented technology, which “enables the graphic representation of nucleic acid sequences and the ability to move them around with ease to create genetic constructs with functional sequences that are precisely arranged.” Essentially, it’s a drag-and-drop function within the company’s free design and assembly software, Gene Designer, for the building blocks of genes.

s012_GeneDesignerSnap211

It may sound trivial, but the feature significantly reduces the time required and errors committed for and during gene construct design for scientists across the globe. Minshull notes:

Even simple editing errors caused by inaccurate in silico sequence manipulations can be challenging to identify and correct without a dedicated design tool. Designing sequences to perform a novel function requires the ability to try out different combinations and make adjustments with ease.

That’s where the problem arose. Both DNA2.0 and Genome Compiler (and other bioCAD developers) allow users to easily reorder a sequence of DNA in a virtual environment. However, Genome Compiler’s use of that feature appeared to blatantly infringe the ‘252 patent.

Source: Genome Compiler.

Source: Genome Compiler.

Perhaps that didn’t matter. Although the USPTO granted DNA2.0 the patent, some questioned whether ‘252 covered patentable material in the first place, invoking parallels from Apple v. Samsung. Was a drag-and-drop feature for nucleic acid sequences a novel or abstract idea? Recent legal cases have successfully challenged the issue of patentable subject matter. Uduak Grace Thomas of BioInform covered the case in January 2013, interviewing Cristopher Holman, a law professor at the University of Missouri:

Clearly, when the patent was granted in 2010, the US Patent and Trademark Office believed that DNA2.0’s patent was valid. However, Holman noted, several recent US Supreme Court decisions — for example Bilski v. Kappos and Mayo Collaborative Services v. Prometheus Laboratories — have raised questions about once-accepted criteria for patentable subject matter.

In Bilski v. Kappos, the Supreme Court rejected a patent for a method of “optimizing a fixed bill system for energy markets” on the grounds that it is an unpatentable abstract idea. And in Mayo Collaborative Services v. Prometheus Laboratories, the Court invalidated Prometheus Labs’ diagnostic patents, ruling that they covered “well-understood, routine, conventional” methods to correlate biological processes with treatment decisions and therefore were not patent eligible.

Previous judicial successes weren’t the only things working in Genome Compiler’s favor (although survivorship bias comes into play here, as there are many more unsuccessful challenges that don’t make the headlines). As it turns out, the user manual for Vector NTI, a sequence design software platform created by InforMax and now owned by Life Technologies, referenced similar ideas to the ‘252 patent, was available for at least one year prior to the patent’s priority review date, and was not mentioned in the ‘252 application. That gave Genome Compiler enough ammunition to challenge DNA2.0, and in December 2013 the startup asked the USPTO to reexamine the ‘252 patent.

Despite the challenge from Genome Compiler it’s important to note that both sides had a valid argument. After all, DNA2.0 was issued a patent, which is generally considered to be valid in a legal setting. The gene synthesis company could also argue that the claims in ‘252 were not abstract, but applicable to a tangible software product. Simply put, the case could have been decided in either party’s favor in a courtroom.

What happened?

A settlement was reached on February 20, 2014 in which Genome Compiler will acquire a worldwide license for ‘252 from DNA2.0. Both companies thought it was an amicable end to a volatile (at least outwardly) situation. From a joint statement issued shortly after the settlement:

Founder and CEO of Genome Compiler, Omri Amirav-Drory: “We are pleased to have arrived at a mutually beneficial business resolution to our dispute.”

Minshull: “Licensing this technology to Genome Compiler benefits both companies and their customers, as well as the industry at large.”

What does the settlement mean?

Handfuls of companies have legitimate shots at capturing market share in emerging technology markets, whether for a biobased chemical, an enzyme utilized in a bioprocess, or the software platforms used in gene design and assembly. However, as we’ve seen in the past, sometimes the race to capturing market share results in overlapping intellectual property. It’s also important to consider who is affected by a dispute. For instance, while Gevo and Butamax continue to wrestle in the judicial system over intellectual property covering the process of making and separating isobutanol from yeast, the outcome is unlikely to affect companies outside of the isobutanol, or even the broader butanol, markets.

On the other hand, patent disputes centered on software platforms that are becoming integrated into the “Design, Build, Test” cycle inherent to the industry could affect a much broader collection of companies. While Genome Compiler and DNA2.0 have resolved their suit involving the ‘252 patent, it’s not immediately clear what the outcome means for related bioCAD platforms that utilize similar user-friendly features. Will DNA2.0 seek licenses from other companies in a similarly amicable (relatively speaking) move? Will the field of bioCAD developers duke it out in court in the future? After all, it’s more difficult to prove novelty for digital ones-and-zeroes (software) than physical real-world products (matter).

I’m not saying design and assembly software is not important for the synthetic biology industry — quite the opposite. Simply put, bioCAD platforms will be incredibly valuable to increasing the productivity and efficiency of designing new organisms while simultaneously slashing costs. However, it seems that bioCAD developers have a higher chance of stepping on each other’s toes with intellectual property covering software. Of course, the debate of whether or not software should be patentable in the first place is nothing new. Some argue it should be, as it increases the value of companies — it certainly adds economic value — but others argue software is nothing more than a thought process, which makes it reliant on algorithms, which makes it math, which anyone can arrive at. There’s good reason many companies choose to keep proprietary software trade secret (as with any other algorithm) rather than submit it to the USPTO.

We’ll get an answer on the broader debate soon enough: This summer the Supreme Court will determine if software is eligible for patent and copyright protection, as currently holds, or just the latter. For the “smaller” big picture view on how the dispute between Genome Compiler and DNA2.0 affects the synthetic biology industry, I’d say the industry can breathe a collective sigh of relief for now. Just remember that this won’t be the last time a patent dispute summons lawyers to the bargaining table, software or otherwise.