The bacterial colonies that are uncoloured have not solved the problem, while ones that are red or green have solved half the problem, and ones that are yellow have solved the whole problem.

Credit: Todd Eckdahl

Related articles

• Alternative to antibiotics under development
• Genome-switching turns one bacterium into another
• Frozen bacterium is 120,000 years old
• New species of bacteria found on human skin
• Quantum computers still work with half their bits missing

SYDNEY: Bacteria can solve complex mathematical problems and may form the building blocks of future supercomputers, according to a new study.

Published in the Journal of Biological Engineering, the proof-of-principle study used glowing bacteria to crack the classic 'Hamilton Path Problem', showing that bacteria can be programmed to do maths.

“Our work demonstrates the potential for using living cells to solve mathematical problems,” said lead researcher Todd Eckdahl, a synthetic biologist at Missouri Western State University in the USA.

Complex mathematical problem

“It supports the view that bacteria can be used to perform computations. Someday, living computers could have applications in medicine, energy, and the environment,” he said.

The Hamilton Path Problem involves working out if a number of connected points on a graph can all be visited, once and only once, using only the existing connections. For example, can you make a car trip visiting Sydney, Melbourne, Alice Springs and Adelaide without leaving the road, or visiting any cities twice?

Although this simple case seems easy to work out, as the number of points increases, the problem becomes much more difficult and time consuming to solve. Current supercomputers would take thousands of years to solve the Hamilton Path Problem with only 20 points, Eckdahl said.

Eckdahl and his team suspected that bacteria might be better at solving this kind of problem than conventional computers. This is because bacteria grow and reproduce very quickly – if you have billions of bacteria working on the same problem in parallel, a solution is likely to evolve, he said.

Coding for colour

Programming the hardware – in this case the bacteria's DNA – is the tricky bit. To see if this was possible, Eckdahl's team programmed some E. Coli cells to solve a simple case using only three points. They modified the bacterial DNA so points were represented as genes which code for colour, with the intervening DNA being the route between points.

These cells were then made to multiply and shuffle their DNA, with different cells representing different solutions to the problem.

Cells which solved half the problem (how to visit two points) glowed either green or red, while cells which had solved the whole problem (how to visit all three cities) glowed both green and red at the same time, causing them to appear yellow. Their DNA was then 'read' to decipher the answer.