Engineering Sequestration-Based\nBiomolecular Classifiers\nwith Shared Resources

Constructing molecular classifiers that enable cells\nto recognize\nlinear and nonlinear input patterns would expand the biocomputational\ncapabilities of engineered cells, thereby unlocking their potential\nin diagnostics and therapeutic applications. While several biomolecular\nclassifier schemes have been designed, the effects of biological constraints\nsuch as resource limitation and competitive binding on the function\nof those classifiers have been left unexplored. Here, we first demonstrate\nthe design of a sigma factor-based perceptron as a molecular classifier\nworking based on the principles of molecular sequestration between\nthe sigma factor and its antisigma molecule. We then investigate how\nthe output of the biomolecular perceptron, i.e., its response pattern\nor decision boundary, is affected by the competitive binding of sigma\nfactors to a pool of shared and limited resources of core RNA polymerase.\nFinally, we reveal the influence of sharing limited resources on multilayer\nperceptron neural networks and outline design principles that enable\nthe construction of nonlinear classifiers using sigma-based biomolecular\nneural networks in the presence of competitive resource-sharing effects.