Scientists from China and the United States have developed an innovative system that allows cells to perform protein-based computations, mimicking the work of artificial neural networks. This new achievement is called “perceptane” – a combination of the words “perceptron” (a key concept in neural networks) and “protein”.
As Day.Az reports with reference to Gazeta.ru, the work was published in the journal Science.
Perceptane is an innovative biological system capable of classifying signals and making decisions. Unlike traditional methods based on DNA and RNA, this approach uses proteins exclusively, making it faster and more efficient.
The system operates thanks to synthetic proteins that form networks capable of processing signals and highlighting only the most significant ones. For example, in one experiment, scientists created a network that, based on protein signals, determined whether a cell should survive or die. To do this, they used special enzymes (proteases), and the results were assessed by the level of fluorescence in the cells.
This technology opens new horizons for the creation of programmed therapies, allowing cells to respond to disease signals. For example, it can ensure selective death of cancer cells, which will significantly improve treatment. In addition, the development could be an important step towards the creation of biological artificial intelligence systems.