Scientists use live bacteria to solve mathematical problems: how the experiment works

The framework of this research –led by the McGill Universityin Canada – is part of the biological computinga discipline that takes advantage of living organisms and processes – from DNA to cultured neurons – to process data. The bet is to explore architectures capable of operating in parallel, with lower energy consumption than electronic systems.

The proposal takes advantage of a natural ability of bacteria to move, reproduce and explore multiple routes. To turn that behavior into a calculation system, scientists designed micro fluid networks that are spliced ​​with tiny interconnected channels. Each path represents a possible solution and the intersections function as logic gates.

Upon entering the network, bacteria pass through these channels and multiply. Each new generation expands the number of biological units involved in the search, allowing numerous alternatives to be explored in parallel. Thus, the system takes advantage of the cell reproduction to increase your calculation capacity without manufacturing new circuits.

The experiment addressed the Subset Sum Problema mathematical challenge that consists of determining if, within a numerical set, there is any combination whose sum is exactly equal to an objective value. The researchers coded the constraints in the network and then identified successful traversals using automated image analysis systems.

He bacterial growth It also requires supervision. Scientists adjust the temperature to control the proliferation of microorganisms and prevent the expansion from ending up saturating the channels. Reading the trajectories allows us to distinguish which of the explored alternatives meet the conditions of the problem.

The challenge is to translate that advantage to larger scale systems. If experts manage to control the growth and distribution of microorganisms, computing platforms could be developed capable of expanding their resources without adding new electronic components.

“Modern electronic computers perform millions of operations per second using a single chip. bacterial divisionwe have a million chips working at the same time. More chips at the same time means that the problem can be solved faster, although each individual processor is slower,” he summarizes. And Nicolauprofessor of Bioengineering and main author of the work.

The researcher also highlighted the potential of this architecture to adapt to problems of different scales. “He exponential growth of bacteria can be adapted to the size of the problem. “It is a completely new way of projecting calculations,” he said.

The project is the result of a scientific collaboration that spanned two decades and involved institutions in Canada, Europe, the United Kingdom and Australia. The results were published in the journal “Microsystems & Nanoengineering” and describe a proof of conceptnot a biological computer ready for commercialization.

Researchers do not see this technology as a substitute for PCs or quantum systems. Its potential would be in specific applications, such as cryptography, molecular biology and some AI developments, where the ability to explore multiple alternatives could provide new tools to solve problems that today demand large processing resources.

Nicolau questioned the idea that AI solves all challenges. “Our approach is designed for problems that computers were not designed to solve, such as mapping the protein structureAccording to the researcher, while the models analyze available data, this technology explores solutions through a physical mechanism.

The potential of biocomputer lies in combining parallel processingcellular reproduction and an energy demand that, according to the authors, could be competitive in specific applications. However, it still needs to overcome obstacles linked to the precision of measurements, the control of bacterial growth and the expansion of microfluidic networks.

The study suggests that, from a certain level of complexity, this approach could outperform some electronic systems on specific tasks. For now, this is a hypothesis. The computing of the future may not only depend on ever smaller transistors, but also on living organisms capable of converting its reproduction into a form of calculation.

Biological computing is not limited to bacteria. Different investigations take advantage of the capabilities of organisms, neurons and molecules to process information through natural mechanisms. Although experimental developments still predominate, these platforms explore alternative forms of calculation that do not depend on conventional electronic circuits.

One of the most striking cases is the Physarum polycephaluma slime mold that lacks a brain, but can find efficient routes between different points. By rearranging its network of internal tubes to distribute nutrients, it solves the spatial optimization. Scientists even tested its integration with electronic devices to take advantage of its adaptability.

Another aspect uses neuronal cultures and brain organoids. Cortical Labsin Australia, and the Universidad Johns Hopkins They investigate how living neurons respond to electrical stimuli and modify their behavior. The «organoid intelligence» seeks to convert these three-dimensional tissues into platforms capable of processing information and learning.

The bacterial spatial computing organizes microorganisms on biochips to process chemical signals based on their location. Researchers of the University of Northwestern They explore how this distribution allows us to configure biological logic circuits that perform functions similar to those of conventional electronics.

The DNA-based computing also advances. Instead of transistors, it uses molecules that store information and carry out operations through chemical reactions. Researchers developed experimental systems to recognize patterns, build molecular neural networks, and perform calculations inspired by Turing machinesdirectly in test tubes.