"DNA-Based Pattern Recognition in Non-Equilibrium Self-Assembly Kinetics"
Originally Published 2 years ago — by Nature.com
Researchers have demonstrated the use of molecular self-assembly for high-dimensional pattern recognition using DNA nanotechnology. By colocalizing molecules in different ways, they were able to control nucleation kinetics and achieve selective assembly of multiple target structures. This work shows that the phase diagram of self-assembling systems can naturally solve complex pattern recognition problems, similar to neural networks. The findings have implications for understanding biological systems and engineering autonomous molecular systems, and could potentially be applied to tasks such as image recognition using existing molecules.