Search

DNA-based neural networks are a type of DNA circuit capable of molecular pattern recognition tasks. Winner-take-all DNA networks have been developed to scale up the complexity of molecular pattern recognition with a simple molecular implementation. This simplicity was achieved by replacing negative weights in individual neurons with lateral inhibition and competition across neurons, eliminating the need for dual-rail representation. Here we introduce a new type of DNA circuit that is called loser-take-all: an output signal is ON if and only if the corresponding input has the smallest analog value among all inputs. We develop a DNA strand-displacement implementation of loser-take-all circuits that is cascadable without dual-rail representation, maintaining the simplicity desired for scalability. We characterize the impact of effective signal concentrations and reaction rates on the circuit performance, and derive solutions for compensating undesired signal loss and rate differences. Using these approaches, we successfully demonstrate a three-input loser-take-all circuit with nine unique input combinations. Complementary to winner-take-all, loser-take-all DNA circuits could be used for recognition of molecular patterns based on their least similarities to a set of memories, allowing classification decisions for patterns that are extremely noisy. Moreover, the design principle of loser-take-all could be more generally applied in other DNA circuit implementations including k-winner-take-all.

Molecular circuits capable of processing temporal information are essential for complex decision making in response to both the presence and history of a molecular environment. A particular type of temporal information that has been recognized to be important is the relative timing of signals. Here we demonstrate the strategy of temporal memory combined with logic computation in DNA strand-displacement circuits capable of making decisions based on specific combinations of inputs as well as their relative timing. The circuit encodes the timing information on inputs in a set of memory strands, which allows for the construction of logic gates that act on current and historical signals. We show that mismatches can be employed to reduce the complexity of circuit design and that shortening specific toeholds can be useful for improving the robustness of circuit behavior. We also show that a detailed model can provide critical insights for guiding certain aspects of experimental investigations that an abstract model cannot. We envision that the design principles explored in this study can be generalized to more complex temporal logic circuits and incorporated into other types of circuit architectures, including DNA-based neural networks, enabling the implementation of timing-dependent learning rules and opening up new opportunities for embedding intelligent behaviors into artificial molecular machines.

Developmental self-assembly of DNA nanostructures provides an ideal platform for studying the power and programmability of kinetically controlled structural growth in engineered molecular systems. Triggered initiation and designated sequencing of assembly and disassembly steps have been demonstrated in structures with branches and loops. Here we introduce a new strategy for selectively activating distinct subroutines in a developmental self-assembly program, allowing structures with distinct properties to be created in response to various molecular signals. We demonstrate this strategy in triggered self-assembly of a DNA ring, the size and growth direction of which are responsive to a key molecule. We articulate that reversible assembly steps with slow kinetics at appropriate locations in a reaction pathway could enable multiple populations of structures with stimulus-responsive properties to be simultaneously created in one developmental program. These results open up a broad design space for the self-assembly of molecules with adaptive behaviors toward advanced control in synthetic materials and molecular motors.

DNA catalysts are fundamental building blocks for diverse molecular information-processing circuits. Allosteric control of DNA catalysts has been developed to activate desired catalytic pathways at desired times. Here we introduce a new type of DNA catalyst that we call a cooperative catalyst: a pair of reversible reactions are employed to drive a catalytic cycle in which two signal species, which can be interpreted as an activator and an input, both exhibit catalytic behavior for output production. We demonstrate the role of a dissociation toehold in controlling the kinetics of the reaction pathway and the significance of a wobble base pair in promoting the robustness of the activator. We show near-complete output production with input and activator concentrations that are 0.1 times the gate concentration. The system involves just a double-stranded gate species and a single-stranded fuel species, as simple as the seesaw DNA catalyst, which has no allosteric control. The simplicity and modularity of the design make the cooperative DNA catalyst an exciting addition to strand-displacement motifs for general-purpose computation and dynamics.

From bacteria following simple chemical gradients to the brain distinguishing complex odour information, the ability to recognize molecular patterns is essential for biological organisms. This type of information-processing function has been implemented using DNA-based neural networks, but has been limited to the recognition of a set of no more than four patterns, each composed of four distinct DNA molecules. Winner-take-all computation has been suggested as a potential strategy for enhancing the capability of DNA-based neural networks. Compared to the linear-threshold circuits and Hopfield networks used previously, winner-take-all circuits are computationally more powerful, allow simpler molecular implementation and are not constrained by the number of patterns and their complexity, so both a large number of simple patterns and a small number of complex patterns can be recognized. Here we report a systematic implementation of winner-take-all neural networks based on DNA-strand-displacement reactions. We use a previously developed seesaw DNA gate motif, extended to include a simple and robust component that facilitates the cooperative hybridization that is involved in the process of selecting a ‘winner’. We show that with this extended seesaw motif DNA-based neural networks can classify patterns into up to nine categories. Each of these patterns consists of 20 distinct DNA molecules chosen from the set of 100 that represents the 100 bits in 10 × 10 patterns, with the 20 DNA molecules selected tracing one of the handwritten digits ‘1’ to ‘9’. The network successfully classified test patterns with up to 30 of the 100 bits flipped relative to the digit patterns ‘remembered’ during training, suggesting that molecular circuits can robustly accomplish the sophisticated task of classifying highly complex and noisy information on the basis of similarity to a memory.

Biochemical circuits made of rationally designed DNA molecules are proofs of concept for embedding control within complex molecular environments. They hold promise for transforming the current technologies in chemistry, biology, medicine and material science by introducing programmable and responsive behaviour to diverse molecular systems. As the transformative power of a technology depends on its accessibility, two main challenges are an automated design process and simple experimental procedures. Here we demonstrate the use of circuit design software, combined with the use of unpurified strands and simplified experimental procedures, for creating a complex DNA strand displacement circuit that consists of 78 distinct species. We develop a systematic procedure for overcoming the challenges involved in using unpurified DNA strands. We also develop a model that takes synthesis errors into consideration and semi-quantitatively reproduces the experimental data. Our methods now enable even novice researchers to successfully design and construct complex DNA strand displacement circuits., DNA circuits hold promise for advancing information-based molecular technologies, yet it is challenging to design and construct them in practice. Thubagere et al. build DNA strand displacement circuits using unpurified strands whose sequences are automatically generated from a user-friendly compiler.

The dynamic interactions between complex molecular structures underlie a wide range of sophisticated behaviors in biological systems. In building artificial molecular machines out of DNA, an outstanding challenge is to develop mechanisms that can control the kinetics of interacting DNA nanostructures and that can compose the interactions together to carry out system-level functions. Here we show a mechanism of DNA tile displacement that follows the principles of toehold binding and branch migration similar to DNA strand displacement, but occurs at a larger scale between interacting DNA origami structures. Utilizing this mechanism, we show controlled reaction kinetics over five orders of magnitude and programmed cascades of reactions in multi-structure systems. Furthermore, we demonstrate the generality of tile displacement for occurring at any location in an array in any order, illustrated as a tic-tac-toe game. Our results suggest that tile displacement is a simple-yet-powerful mechanism that opens up the possibility for complex structural components in artificial molecular machines to undergo information-based reconfiguration in response to their environments.

Self-assembled DNA nanostructures enable nanometre-precise patterning that can be used to create programmable molecular machines and arrays of functional materials. DNA origami is particularly versatile in this context because each DNA strand in the origami nanostructure occupies a unique position and can serve as a uniquely addressable pixel. However, the scale of such structures has been limited to about 0.05 square micrometres, hindering applications that demand a larger layout and integration with more conventional patterning methods. Hierarchical multistage assembly of simple sets of tiles can in principle overcome this limitation, but so far has not been sufficiently robust to enable successful implementation of larger structures using DNA origami tiles. Here we show that by using simple local assembly rules that are modified and applied recursively throughout a hierarchical, multistage assembly process, a small and constant set of unique DNA strands can be used to create DNA origami arrays of increasing size and with arbitrary patterns. We illustrate this method, which we term ‘fractal assembly’, by producing DNA origami arrays with sizes of up to 0.5 square micrometres and with up to 8,704 pixels, allowing us to render images such as the Mona Lisa and a rooster. We find that self-assembly of the tiles into arrays is unaffected by changes in surface patterns on the tiles, and that the yield of the fractal assembly process corresponds to about 0.95^(m − 1) for arrays containing m tiles. When used in conjunction with a software tool that we developed that converts an arbitrary pattern into DNA sequences and experimental protocols, our assembly method is readily accessible and will facilitate the construction of sophisticated materials and devices with sizes similar to that of a bacterium using DNA nanostructures.

Scaling up the complexity and diversity of synthetic molecular structures will require strategies that exploit the inherent stochasticity of molecular systems in a controlled fashion. Here we demonstrate a framework for programming random DNA tilings and show how to control the properties of global patterns through simple, local rules. We constructed three general forms of planar network—random loops, mazes and trees—on the surface of self-assembled DNA origami arrays on the micrometre scale with nanometre resolution. Using simple molecular building blocks and robust experimental conditions, we demonstrate control of a wide range of properties of the random networks, including the branching rules, the growth directions, the proximity between adjacent networks and the size distribution. Much as combinatorial approaches for generating random one-dimensional chains of polymers have been used to revolutionize chemical synthesis and the selection of functional nucleic acids, our strategy extends these principles to random two-dimensional networks of molecules and creates new opportunities for fabricating more complex molecular devices that are organized by DNA nanostructures. Programming stochastic self-assembly of DNA origami tiles to create complex patterns with controlled properties.

A solution-phase DNA chip is produced using DNA origami-based asymmetric self-assembled tiles incorporating single-stranded DNA linear probes. This DNA origami chip is fully spatially addressable, without having to use position index oligonucleotides. The DNA hybridization is readily detected via atomic force microscopy.

This paper presents a DNA algorithm which adds two nonnegative binary integers using self-assembly in constant steps. The approach has the benefit of greater experimental simplicity when compared with previous DNA addition algorithms. For the addition of two binary n-bit integers, O(n) is different from DNA strands and only O(1) biochemical experimental procedures are required.

The prospects of programming molecular systems to perform complex autonomous tasks have motivated research into the design of synthetic biochemical circuits. Of particular interest to us are cell-free nucleic acid systems that exploit non-covalent hybridization and strand displacement reactions to create cascades that implement digital and analogue circuits. To date, circuits involving at most tens of gates have been demonstrated experimentally. Here, we propose a simple DNA gate architecture that appears suitable for practical synthesis of large-scale circuits involving possibly thousands of gates.

To construct sophisticated biochemical circuits from scratch, one needs to understand how simple the building blocks can be and how robustly such circuits can scale up. Using a simple DNA reaction mechanism based on a reversible strand displacement process, we experimentally demonstrated several digital logic circuits, culminating in a four-bit square-root circuit that comprises 130 DNA strands. These multilayer circuits include thresholding and catalysis within every logical operation to perform digital signal restoration, which enables fast and reliable function in large circuits with roughly constant switching time and linear signal propagation delays. The design naturally incorporates other crucial elements for large-scale circuitry, such as general debugging tools, parallel circuit preparation, and an abstraction hierarchy supported by an automated circuit compiler.

The impressive capabilities of the mammalian brain--ranging from perception, pattern recognition and memory formation to decision making and motor activity control--have inspired their re-creation in a wide range of artificial intelligence systems for applications such as face recognition, anomaly detection, medical diagnosis and robotic vehicle control. Yet before neuron-based brains evolved, complex biomolecular circuits provided individual cells with the 'intelligent' behaviour required for survival. However, the study of how molecules can 'think' has not produced an equal variety of computational models and applications of artificial chemical systems. Although biomolecular systems have been hypothesized to carry out neural-network-like computations in vivo and the synthesis of artificial chemical analogues has been proposed theoretically, experimental work has so far fallen short of fully implementing even a single neuron. Here, building on the richness of DNA computing and strand displacement circuitry, we show how molecular systems can exhibit autonomous brain-like behaviours. Using a simple DNA gate architecture that allows experimental scale-up of multilayer digital circuits, we systematically transform arbitrary linear threshold circuits (an artificial neural network model) into DNA strand displacement cascades that function as small neural networks. Our approach even allows us to implement a Hopfield associative memory with four fully connected artificial neurons that, after training in silico, remembers four single-stranded DNA patterns and recalls the most similar one when presented with an incomplete pattern. Our results suggest that DNA strand displacement cascades could be used to endow autonomous chemical systems with the capability of recognizing patterns of molecular events, making decisions and responding to the environment.

The prospects of programming molecular systems to perform complex autonomous tasks has motivated research into the design of synthetic biochemical circuits. Of particular interest to us are cell-free nucleic acid systems that exploit non-covalent hybridization and strand displacement reactions to create cascades that implement digital and analog circuits. To date, circuits involving at most tens of gates have been demonstrated experimentally. Here, we propose a DNA catalytic gate architecture that appears suitable for practical synthesis of large-scale circuits involving possibly thousands of gates.