Single-Stranded Architectures for Computing

RNA is a chain of ribonucleotides of four kinds (denoted respectively by the letters A, C, G, U). While being synthesized sequentially from its template DNA (transcription), it folds upon itself into intricate higher-dimensional structures in such a way that the free energy is minimized, that is, the more hydrogen bonds between ribonucletoides or larger entropy a structure has, the more likely it is chosen, and furthermore the minimization is done locally. This phenomenon is called cotranscriptional folding (CF). It has turned out to play significant roles in in-vivo computation throughout experiments and recently proven even programmable artificially so as to self-assemble a specific RNA rectangular tile structure in vitro. The next step is to program a computation onto DNA in such a way that the computation can be “called” by cotranscriptional folding. In this novel paradigm of computation, what programmers could do is only twofold: designing a template DNA and setting environmental parameters. Oritatami is an introductory “toy” model to this paradigm of computation. In this model, programmars are also allowed to employ an arbitrarily large finite alphabet \(\varSigma \) as well as an arbitrarily complex rule set for binding over \(\varSigma \times \varSigma \). We shall present known architectures of computing in the oritatami model from a simple half-adder to Turing machine along with several programming techniques of use, with hope that they will inspire in-vivo architectures of CF-driven self-assemblable computers, which could be even heritable.