Probabilistic reasoning with a Bayesian DNA device based on strand displacement

Abstract

Abstract. We present a computing model based on the DNA strand displacement technique which performs Bayesian inference. The model will take single stranded DNA as input data, representing the presence or absence of a specific molecular signal (evidence). The program logic encodes the prior probability of a disease and the conditional probability of a signal given the disease playing with a set of different DNA complexes and their ratios. When the input and program molecules interact, they release a different pair of single stranded DNA species whose relative proportion represents the application of Bayes' Law: the conditional probability of the disease given the signal. The models presented in this paper can empower the application of probabilistic reasoning in genetic diagnosis in vitro.

Probabilistic Reasoning with an Enzyme-Driven DNA Device

“…We presented a similar model in [24], which used DNA strand displacement instead of Rondelez's DNA toolbox. The rest of the chapter is structured as follows.…”

“…This research has addressed the two main improvement opportunities of the work that we presented elsewhere [24]:…”

“…Signal attenuation. In theory, the model in [24] was also able to deal with multiple input signals by cascading the outputs as inputs of other conditional probability devices downstream. However, each inference iteration would attenuate the signal by an average of 50%.…”

Operant conditioning of stochastic chemical reaction networks

“…In this work, we implement probabilistic decision-making in an abstract CRN, producing a system that can make probabilistic decisions in response to a stimulus at the population level. This is novel because much previous work in this direction has focused either on probabilistic decision-making at the single molecule level [14], which indirectly produces deterministic behavior at the population level, as shown in Fig 1B , or on computing probabilities without acting on them in a probabilistic manner [15,16]. However, in our scheme a population of 100 individual molecules with 70:30 probabilities programmed for a decision between two possible responses X and Y to a stimulus, will either transition to 100 copies of the response X molecule with 70% probability or to 100 copies of the response Y molecule with 30% probability, as shown in Fig 1C, as opposed to simply producing a 70:30 split.…”

Inference Models in DNA Computing

“…A preliminary version of the model presented in this chapter was selected for oral presentation in the 18th International Conference on DNA Computing and Molecular Programming and published in the LNCS proceedings (Sainz de Murieta and Rodríguez-Patón, 2012c). This chapter builds on that work, presenting improvements such as a kinetic model and the results of an in silico simulation.…”