TR: Another outcome of the discussions is the formation of an industry organization for DNA synthesis companies. What is the vision for this organization?
DE: Everyone running a DNA synthesis company has the same problem. You need to check what you’re making to ensure you’re not unknowingly producing something that could cause harm to either someone in the factory, the person you’re delivering it to, or someone else. The solution to this problem can be shared. And if you share the investment in the technology, you’ll be able to develop better technology than anyone could individually. Effective cooperation and industry leadership are needed to make it work.
TR: What is your role in the nascent organization?
DE: We’re trying to help get the organization off the ground by organizing political support – I think it’s irresponsible for the academic research community to call for something to happen without helping. The organization itself will be independent of academia.
TR: During the recent Synthetic Biology meeting, someone from one of the DNA synthesis companies described how hard it was to get the government’s help in vetting orders. He said he’d received an order containing a potentially dangerous sequence and called several government agencies, trying to determine how to proceed. No one could tell him what to do.
DE: This has been a real challenge in the U.S. for six years. Nobody has been able to identify a point-of-contact within government who could help. It’s one of the obvious reasons for starting an industry group, so that the group provides a centralized point-of-contact for government interaction and, as appropriate, oversight.
TR: A large part of the industry organization’s role will be to guide development of new software. How does the current screening software work?
DE: One example is a piece of software called BlackWatch, developed by Rob Jones at Craic Computing. BlackWatch is a computer program that checks to make sure you’re not unknowingly making a sequence of DNA from the select agent list (a list of pathogens).
TR: What are the limitations of the BlackWatch software?
DE: One limitation is the high false positive rate. If you’re comparing the requested sequence against a set of sequences that includes the entire genome of a pathogenic bacterium, many of those genes look like genes of E. coli [bacteria commonly used in research]. Sorting out what’s actually versus apparently dangerous is slow and expensive. It often requires a PhD to manually make the decision. A second limitation is the fact that it’s naive to ask a computer program looking at DNA sequences to infer the intent of the designer of the DNA sequence. Instead, you want the software nested in a decision-making process. Who is ordering the DNA and where is it being shipped? As important, by asking these questions it helps to ensure that the people designing the DNA and their local community are paying attention to issues of biological safety and security.
TR: Any idea what the new software will look like?
DE: There are a number of basic questions that need to be explored. What makes a sequence of DNA dangerous? Can you detect something natural versus engineered? It’s a research process, so there are lots of ideas on how to go about it.
TR: Some people say the biggest potential comes from clandestine government programs, which would actually have the resources to work with these materials. Will the outlined declaration prevent the latter possibility?
DE: No, it doesn’t directly address the issue that some government might start a biological weapons program. But, indirectly, if we don’t have a vibrant and technically superior open research community, the possibility of future harm from clandestine state programs will certainly increase.