Last month, I described how companies that make artificial DNA were locked in a struggle over how much effort they should spend to keep their products out of the hands of terrorists. At the time, a leading industry group, the International Association of Synthetic Biology (IASB), had just adopted a formal Code of Conduct that required members to, inter alia, find out what requested genes actually did before selling them to customers. This process required significant human judgment and was therefore expensive.
Meanwhile, two other companies – DNA2.0 and Geneart – had proposed a different method in which incoming threats would be compared against a predefined threat list. Given the limits of current technology, DNA2.0/Geneart clearly knew that their system was less effective than IASB’s proposal. Instead, its main advantage was that it could be readily automated. This made it, as its proponents liked to point out, “fast” and “cheap.”
At the time I wrote, the two sides had just staked out their respective positions. In the long run, it was clear that only one standard could prevail. But which would it be? Clearly, a Silicon Valley-style “standards war” was brewing.
What a difference a month makes. Readers familiar with regulation and diplomacy know that government action can take a long time. Private sector “standards wars” move much, much faster.
On November 17, DNA2.0 and Geneart resurfaced as part of a new International Gene Synthesis Consortium (IGSC) with fellow gene synthesis providers Genscript, Blue Heron, and IDT. Predictably, the IGSC’s first official act was to announce its own Harmonized Screening Protocol.
Surprisingly, the new document echoed IASB’s earlier code in practically every detail. This included abandoning the “fast” and “cheap” approach that two of its members had previously advocated. Instead, the Protocol joined the IASB Code in requiring companies to a) compare incoming customer orders against all of Genbank (the U.S. government’s exhaustive database of known genes), and b) use human experts to investigate all close matches to pathogen or toxin sequences. This convergence between the Protocol and Code was tremendously encouraging. Whichever standard prevailed, it looked like the gene synthesis industry was converging on a single (and high) standard of practice.
That said, there was a fly in the ointment. IASB had developed its code in open meetings – indeed, representatives of Geneart, Blue Heron, and Nature had all been present when IASB finalized its Code on November 3. By contrast, IGSC wrote its Protocol behind in secret behind closed doors.
IGSC justified this procedure on the ground that its members collectively owned 80 percent of worldwide gene synthesis capacity. As IDT’s Damon Terrill told one reporter, “We are the gene synthesis industry.” Or as Louis XIV might have put it, “L'état, c'est moi.”
The ethical argument behind such statements is obscure. After all, few people accept large market share as proof of virtue or even wisdom. Why should outsiders – including dozens of gene synthesis companies and the general public – trust standards developed in secret by a five-member club?
IASB’s Code was credible precisely because the Cambridge meeting was open to IGSC members and even a reporter from Nature. Under these circumstances, it would have been difficult or impossible for members to sweep flaws under the rug. Whatever its flaws, it is hard to doubt the Code’s sincerity.
IGSC members, on the other hand, have repeatedly suggested that while they admire what IASB is doing, they also have a different “perspective about the scale of the gene-synthesis industry, which helps us to decide what are practically implementable decisions,” according to Nature News. This sounds like bad economics. If small companies can endorse IASB’s standard despite their comparatively thin profit margins, then the IGSC companies – all of which enjoy large economies of scale – surely can.
In an open meeting, this sort of issue would have been raised and – giving IGSC the benefit of the doubt – addressed immediately. Instead, it has been left to fester. Nor does IGSC say what (if any) parts of the Protocol have been shaped by its members’ large company “perspective.” Smaller companies and consumers will rightly wonder about this.
In the normal course of things, one might have expected the marketplace to resolve the “standards war” between IASB’s Code and IGSC’s Protocol in short order. After all, gene synthesis companies will eventually have to adopt one standard or the other in order to assure their customers (notably including Big Pharma) that they take screening seriously. Once a critical mass of gene synthesis companies endorses the Code (or Protocol), the less popular standard will go the way of Betamax tapes.
One might have expected the U.S. government to be delighted by the prospect of two industry groups fighting over which (high) standard to adopt. If so, it would have been natural for the government to draft a document that endorsed the Code’s and/or Protocol’s best features. Strangely, this did not happen.
Instead, the federal government entered the fray on November 27 by writing “Screening Framework Guidance for Synthetic Double Stranded DNA Providers” which proposed an entirely new procedure called Best Match. The idea was that companies should compare customer orders against (a) all of Genbank and (b) the U.S. government’s so-called Select Agent list of known weapons pathogens and toxins. If the order’s Best Match did not involve a Select Agent, no further human follow-up would be required.
Alert readers will already have recognized that Best Match is just another version of DNA2.0 and Geneart’s earlier, “predefined list” suggestion. (Indeed, the government’s Guidance was probably written in this time frame.) For this reason, it has the same virtues (fast, cheap) but also the same vice (incompleteness) as every other predefined list scheme. Prior to the Guidance, the market had rejected this tradeoff.
Now, suddenly, the Guidance has reopened the question. Indeed, Nature News reported on December 4 that some IGSC members were already praising Best Match as an acceptable alternative.
The good news, of course, is that the Guidance is still only in draft form. In principle, it would be easy to fix it by, for example, requiring companies to augment Best Match searches with a second, parallel procedure in which human experts must examine close matches to any Genbank pathogen or toxin sequence. Beyond this, the federal government should also consider reminding industry that standards should be set openly.
Readers do not have to take my word for any of this. The IASB Code, IGSC Protocol, and federal draft Guidance are all short documents. Readers can compare for themselves. And it is important to do this.
The current “standards war” between IASB, IGSC and the federal government will likely set screening policy for a generation. Policy professionals – including most readers of this blog – have an obligation to speak up. Interested readers should e-mail comments to email@example.com on or before January 26.