Evolutionary chemical learning in dimerization networks
This work advances adaptive, energy-efficient molecular computing systems by bridging synthetic biology and machine learning, though it is incremental as it builds on existing directed evolution and analog computation concepts.
The authors tackled the problem of implementing complex learning tasks like multiclass classification using chemical systems, by developing Competitive Dimerization Networks (CDNs) trained through directed evolution, resulting in classifiers with strong output contrast and high mutual information, closely matching in silico gradient descent performance.
We present a novel framework for chemical learning based on Competitive Dimerization Networks (CDNs) - systems in which multiple molecular species, e.g. proteins or DNA/RNA oligomers, reversibly bind to form dimers. We show that these networks can be trained in vitro through directed evolution, enabling the implementation of complex learning tasks such as multiclass classification without digital hardware or explicit parameter tuning. Each molecular species functions analogously to a neuron, with binding affinities acting as tunable synaptic weights. A training protocol involving mutation, selection, and amplification of DNA-based components allows CDNs to robustly discriminate among noisy input patterns. The resulting classifiers exhibit strong output contrast and high mutual information between input and output, especially when guided by a contrast-enhancing loss function. Comparative analysis with in silico gradient descent training reveals closely correlated performance. These results establish CDNs as a promising platform for analog physical computation, bridging synthetic biology and machine learning, and advancing the development of adaptive, energy-efficient molecular computing systems.