Scaling up molecular pattern recognition with DNA-based winner-take-all neural networks

From bacteria following simple chemical gradients1 to the brain distinguishing complex odour information2, the ability to recognize molecular patterns is essential for biological organisms. This type of information-processing function has been implemented using DNA-based neural networks3, 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 computation4 has been suggested5,6 as a potential strategy for enhancing the capability of DNA-based neural networks. Compared to the linear-threshold circuits7 and Hopfield networks8 used previously3, winner-take-all circuits are computationally more powerful4, 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-displacement9,10 reactions. We use a previously developed seesaw DNA gate motif3,11,12, extended to include a simple and robust component that facilitates the cooperative hybridization13 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.DNA-strand-displacement reactions are used to implement a neural network that can distinguish complex and noisy molecular patterns from a set of nine possibilities—an improvement on previous demonstrations that distinguished only four simple patterns.

[1]  Teruo Fujii,et al.  Scaling down DNA circuits with competitive neural networks , 2013, Journal of The Royal Society Interface.

[2]  G. Seelig,et al.  Dynamic DNA nanotechnology using strand-displacement reactions. , 2011, Nature chemistry.

[3]  Bernard Yurke,et al.  Using DNA to Power Nanostructures , 2003, Genetic Programming and Evolvable Machines.

[4]  P. Redgrave,et al.  The basal ganglia: a vertebrate solution to the selection problem? , 1999, Neuroscience.

[5]  Lulu Qian,et al.  Compiler-aided systematic construction of large-scale DNA strand displacement circuits using unpurified components , 2017, Nature Communications.

[6]  John J. Hopfield,et al.  Neural networks and physical systems with emergent collective computational abilities , 1999 .

[7]  Conrad Steenberg,et al.  NUPACK: Analysis and design of nucleic acid systems , 2011, J. Comput. Chem..

[8]  Georg Seelig,et al.  A DNA Neural Network Constructed from Molecular Variable Gain Amplifiers , 2017, DNA.

[9]  Wolfgang Maass,et al.  On the Computational Power of Winner-Take-All , 2000, Neural Computation.

[10]  Xi Chen,et al.  Rational, modular adaptation of enzyme-free DNA circuits to multiple detection methods , 2011, Nucleic acids research.

[11]  D. Y. Zhang,et al.  Control of DNA strand displacement kinetics using toehold exchange. , 2009, Journal of the American Chemical Society.

[12]  Georg Seelig,et al.  DNA-Based Fixed Gain Amplifiers and Linear Classifier Circuits , 2010, DNA.

[13]  Lulu Qian,et al.  Supporting Online Material Materials and Methods Figs. S1 to S6 Tables S1 to S4 References and Notes Scaling up Digital Circuit Computation with Dna Strand Displacement Cascades , 2022 .

[14]  Chrisantha Fernando,et al.  Molecular circuits for associative learning in single-celled organisms , 2008, Journal of The Royal Society Interface.

[15]  Luca Cardelli,et al.  Programmable chemical controllers made from DNA. , 2013, Nature nanotechnology.

[16]  D. Stefanovic,et al.  Training a molecular automaton to play a game. , 2010, Nature nanotechnology.

[17]  A. Turberfield,et al.  A DNA-fuelled molecular machine made of DNA , 2022 .

[18]  David Yu Zhang,et al.  Cooperative hybridization of oligonucleotides. , 2011, Journal of the American Chemical Society.

[19]  Eun Jeong Cho,et al.  Applications of aptamers as sensors. , 2009, Annual review of analytical chemistry.

[20]  Luca Cardelli,et al.  The Cell Cycle Switch Computes Approximate Majority , 2012, Scientific Reports.

[21]  Masami Hagiya,et al.  Evolving cheating DNA networks: a Case Study with the Rock–Paper–Scissors Game , 2013, Advances in Artificial Life, ECAL 2013.

[22]  Matthew R. Lakin,et al.  Design of a biochemical circuit motif for learning linear functions , 2014, Journal of The Royal Society Interface.

[23]  Saburo Muroga,et al.  Threshold logic and its applications , 1971 .

[24]  K. Mori,et al.  The olfactory bulb: coding and processing of odor molecule information. , 1999, Science.

[25]  G. Wadhams,et al.  Making sense of it all: bacterial chemotaxis , 2004, Nature Reviews Molecular Cell Biology.

[26]  J. Stock,et al.  Bacterial chemotaxis , 2003, Current Biology.

[27]  Raúl Rojas,et al.  Neural Networks - A Systematic Introduction , 1996 .

[28]  L. Deng,et al.  The MNIST Database of Handwritten Digit Images for Machine Learning Research [Best of the Web] , 2012, IEEE Signal Processing Magazine.

[29]  Jehoshua Bruck,et al.  Neural network computation with DNA strand displacement cascades , 2011, Nature.

[30]  Erik Winfree,et al.  Neural Network Computation by In Vitro Transcriptional Circuits , 2004, NIPS.