A McGill-led research team has developed a bacteria-based biocomputer that can solve large, complex problems using a small fraction of the memory and resources required by standard ("solid-state") computers. Possible applications include encryption, molecular biology and artificial intelligence (AI), the researchers said.
Solid-state computers test each possible solution sequentially, which can take years or even centuries to deliver results for complex problems. In contrast, bacteria can explore many possible solutions simultaneously. Additionally, bacteria can divide to create more biological "processors," they explained.
"Modern electronic computers make millions of operations per second using one chip," said Dan Nicolau , senior author and Professor of Bioengineering. "With bacterial division, we have one million chips working at the same time. More chips mean you can resolve the problem faster, even if for each "processor," the number of operations per second is slower.
"The exponential division of bacteria can go to the size of the problem. It's an entirely new way of computing," he said.
Exploring possibilities in parallel
The project is the result of a 20-year, cross-institutional collaboration among researchers in Canada, Europe, the United Kingdom and Australia.
For their biocomputer, the researchers translated a classic problem into a graph of possible choices, then recreated it as a network of tiny, interconnected channels through which bacteria could explore many routes. Each route through the network represented a different possible solution, while special "logic gate" junctions represented decision points.
As rapidly dividing bacteria moved through the channels, they collectively explored many routes at once, testing large numbers of possible solutions in parallel. To prevent overcrowding, the researchers can adjust the temperature of the computer to control bacterial growth.
By observing where the bacteria travelled and using automated image analysis to trace successful routes through the network, the researchers were able to identify combinations that satisfied the problem's constraints.
A resource-efficient tool
The biocomputer is not meant to replace classical electronic or quantum computers, but rather to provide a resource-efficient, cost-effective tool through which to solve previously unsolvable problems, the researchers said.
"Our way is tailored for problems that electronic computers are not designed to solve, such as mapping the structure of proteins," Nicolau said. "It was reported that AI figured this out, but this is only half-true, because it can only scan the available data. We propose this technology to solve problems, not estimate based on past information."
About the study
" Amplification of computational power by the multiplication of bacteria exploring microfluidic networks encoding mathematical problem ," by Ayyappasamy Sudalaiyadum Perumal, Falco C. M. J. M. van Delft, Giulia Ippoliti, Ondřej Kašpar, Viola Tokárová, Monalisha Nayak, Matthew Cho, Jessica Li, Anja van Langen-Suurling, Charles de Boer, Frank Dirne, Dan V. Nicolau Jr and Dan V. Nicolau, was published in Nature Microengineering and Nanosystems.
The study was funded by the (U.S.) Defense Advanced Research Projects Agency, the Natural Sciences and Engineering Research Council of Canada, the Social Sciences and Humanities Research Council of Canada, an Australian Future Fellowship and the European Union.
