Software Is Designing Useful Microbes That Don't Exist in Nature
A startup called 20n has already made a Tylenol clone with its genetically-engineered microbes.
A map of every chemical that can be made biologically, as discovered by 20n's algorithms. Image: 20n
A new biotech startup called 20n has developed software that helps design genetically-engineered microbes to organically produce chemicals that would otherwise have to be made through more complicated industrial processes.
The company, which launched in February and presented at startup incubator Y Combinator's Demo Day on Wednesday, was co-founded by UC Berkeley professor J. Christopher Anderson and post-doc Saurabh Srivastava, and funded through multiple contracts with DARPA. It uses a computer-based biochemistry model to simulate how a cell will behave when foreign DNA is added.
"There are lots of DNA fragments that you can find across all species of the world, and being able to effectively say what will happen when these are added is unprecedented," Srivastava told Motherboard by email. "The molecules our software-designed organisms can produce aren't even constrained to naturally-occurring molecules."
To explain its microbe engineering, 20n uses the analogy of yeast producing alcohol when beer is brewed—the goal is to have software-designed and genetically-engineered microbes produce chemicals that would normally need to be produced through industrial processes. The first successful implementation of this technology is bacteria they engineered to produce paracetamol, the active ingredient in Tylenol.
"Our software predicted that adding only a single gene, from the common grocery store mushroom, would enable E. coli to produce this common headache medicine," Srivastava said. "When we tried it, it worked."
Srivastava and Anderson began developing the software in 2011 after they met at Berkeley. Anderson had been working in a wetlab engineering microbes for 17 years and Srivastava was a postdoc with a Ph.D in computer science.
"With such different backgrounds, we spent an entire year simply developing a common vocabulary to be able to discuss this challenging interface," Srivastava said.
According to Srivastava, 20n's most distinguishing characteristic is that it is a pure software entity. He said that experiments done by 20n are entirely automated and done in contract research organizations (CROs) or labs in the cloud.
"Our core expertise is on the design of the DNA that goes into the microbes, and other organizations build that DNA, get it into the cells, grow them up, and see what chemicals get produced," he said.
20n is partnering with various entities to more quickly bring the technology to market. Soon, Srivastava said, chemicals across the spectrum of organic compounds could be created with this technology, including therapeutics, materials, flavors, fragrances and more.