The DNA computer described by researchers can carry out calculations using billions of molecules suspended in a tiny drop of water. The team says the system leverages physical principles to deliver a form of computation that is more energy efficient than conventional electronic hardware.
Instead of consuming power to force operations through stepwise processes, the setup is engineered so that its most energetically favorable state corresponds to the right answer. According to the researchers, this means it can use less energy than other biological computing approaches that integrate living cells and traditional components.
“The clever part is that the binding process is competitive,” said Damien Woods of Maynooth University, a co-author of the study. Molecules compete to bind to a scaffold, he said, and the system settles into the lowest energy state that encodes the result.
The team reports the Scaffolded DNA Computer, or SDC, in a paper published Sept. 16 in Nature. They evaluated 10 programs, including 100-bit tasks. Simple arithmetic such as 10 + 3 took about 30 seconds for the SDC to resolve.
The researchers emphasized that potential uses remain speculative. Over the long term, DNA-based platforms could aid molecular data storage, enable energy-efficient computation, or operate within living cells.
“Molecular computers like this are not trying to replace electronic ones,” said co-author Abeer Eshra. “They could be used in biological environments, smart materials and archival DNA data storage,” adding that the approach could offer built-in error correction.
A computer made from DNA
The SDC consists of short DNA strands that interact with a longer DNA scaffold. The mixture is placed in salt water and cycled through heating and cooling.
As strands interact, they assemble according to programmed rules that define the computation. The sequences determine which pieces attach to each other and to positions on the scaffold. By designing these binding rules, the researchers effectively program the calculation.
The method exploits thermodynamics, the natural drift of systems toward lower energy states. During heating and cooling, strands compete to form the most stable arrangement. The favored configuration encodes the answer, so the molecules “compute” by interacting until the stable state emerges.
Although each SDC reaction is tiny, the number of molecules involved is vast.
“A small droplet of liquid contains billions, and sometimes trillions, of DNA strands,” Eshra said in a statement. “These strands interact with one another to produce a result.”
A reusable molecular computer
Across their experiments, the researchers demonstrated more than 700 computations. Programs included addition, multiplication by 3, division by 2, and eight-bit parity detection used for error checking. Small problems finished in under a minute, which is notable given the required chemical reactions.
“They’re trivial calculations you could easily do faster yourself,” said co-author Constantine Evans. “Our system uses just a handful of molecules, never really following an organized process of steps, never making irreversible steps, and yet ending up with the right answer.”
More complex computations took longer. A larger sum, in a range between roughly 11 million and 34 million, required up to 14 hours.
The team said the platform is programmable and reusable. Unlike many prior molecular computers built for one-time demonstrations, the SDC can be run repeatedly without special fuels or tightly timed reactions.
“Many molecular computers to date relied on specially prepared components, or molecular fuels, or carefully timed reactions,” Eshra said. “Our DNA computer instead works by throwing the molecules together and letting it relax towards equilibrium.”
Three programs were successfully repeated up to 24 times. In one case, the researchers reran an experiment 1.5 years later after the sample had partially dried. Adding water allowed the calculation to run again.
For now, the work primarily shows that thermodynamic principles can carry out useful computation. The authors said future efforts include designing improved scaffolds, enhancing readouts, and exploring applications in DNA data storage.