Feedforward Chemical Neural Network : An In Silico Chemical System

Inspired by natural biochemicals that perform complex information processing within living cells, we design and simulate a chemically implemented feedforward neural network, which learns by a novel chemical-reaction-based analogue of backpropagation. Our network is implemented in a simulated chemical system, where individual neurons are separated from each other by semipermeable cell-like membranes. Our compartmentalized, modular design allows a variety of network topologies to be constructed from the same building blocks. This brings us towards general-purpose, adaptive learning in chemico: wet machine learning in an embodied dynamical system.

[1]  Kazuto Tominaga,et al.  Modeling Molecular Computing Systems by an Artificial ChemistryIts Expressive Power and Application , 2007, Artificial Life.

[2]  R. Jackson,et al.  General mass action kinetics , 1972 .

[3]  Arto Salomaa,et al.  Algorithmic Bioprocesses , 2009, Natural Computing Series.

[4]  G. Seelig,et al.  DNA as a universal substrate for chemical kinetics , 2010, Proceedings of the National Academy of Sciences.

[5]  Marvin Minsky,et al.  Perceptrons: An Introduction to Computational Geometry , 1969 .

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

[7]  Matthew R. Lakin,et al.  Supervised Learning in Adaptive DNA Strand Displacement Networks. , 2016, ACS synthetic biology.

[8]  Robert Langer,et al.  Small-scale systems for in vivo drug delivery , 2003, Nature Biotechnology.

[9]  Nicolas E. Buchler,et al.  On schemes of combinatorial transcription logic , 2003, Proceedings of the National Academy of Sciences of the United States of America.

[10]  Ricard V. Solé,et al.  Life Cycle of a Minimal ProtocellA Dissipative Particle Dynamics Study , 2007, Artificial Life.

[11]  F. Kleinhans,et al.  Membrane permeability modeling: Kedem-Katchalsky vs a two-parameter formalism. , 1998, Cryobiology.

[12]  Robert A. Copeland,et al.  Enzymes: A Practical Introduction to Structure, Mechanism, and Data Analysis , 1996 .

[13]  Gheorghe Paun,et al.  A guide to membrane computing , 2002, Theor. Comput. Sci..

[14]  A. Kargol,et al.  Mechanistic formalism for membrane transport generated by osmotic and mechanical pressure. , 2003, General physiology and biophysics.

[15]  Gheorghe Paun,et al.  Computing with Membranes , 2000, J. Comput. Syst. Sci..

[16]  S. Basu,et al.  A synthetic multicellular system for programmed pattern formation , 2005, Nature.

[17]  Ben Shirt-Ediss.  Dynamical Systems Analysis of a Protocell Lipid Compartment , 2009, ECAL.

[18]  Christof Teuscher,et al.  An Analog Chemical Circuit with Parallel-Accessible Delay Line for Learning Temporal Tasks , 2014, ALIFE.

[19]  Keshab K. Parhi,et al.  Digital Signal Processing With Molecular Reactions , 2012, IEEE Design & Test of Computers.

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

[21]  M. Hénon,et al.  A two-dimensional mapping with a strange attractor , 1976 .

[22]  Geoffrey E. Hinton,et al.  Learning representations by back-propagating errors , 1986, Nature.

[23]  M. Okamoto,et al.  Switching mechanism of a cyclic enzyme system: role as a "chemical diode". , 1987, Bio Systems.

[24]  Christof Teuscher,et al.  Training an asymmetric signal perceptron through reinforcement in an artificial chemistry , 2014, Journal of The Royal Society Interface.

[25]  Jie-Hong Roland Jiang,et al.  Building reconfigurable circuitry in a biochemical world , 2014, 2014 IEEE Biomedical Circuits and Systems Conference (BioCAS) Proceedings.

[26]  R. Weiss,et al.  Multi-input Rnai-based Logic Circuit for Identification of Specific , 2022 .

[27]  Kurt Hornik,et al.  Multilayer feedforward networks are universal approximators , 1989, Neural Networks.

[28]  Peter Dittrich,et al.  Chemical Computing , 2004, UPP.

[29]  Christopher A. Voigt,et al.  Robust multicellular computing using genetically encoded NOR gates and chemical ‘wires’ , 2011, Nature.

[30]  Christof Teuscher,et al.  Delay Line as a Chemical Reaction Network , 2014, Parallel Process. Lett..

[31]  H. Maturana,et al.  Autopoiesis and Cognition : The Realization of the Living (Boston Studies in the Philosophy of Scie , 1980 .

[32]  Péter Érdi,et al.  Mathematical Models of Chemical Reactions: Theory and Applications of Deterministic and Stochastic Models , 1989 .

[33]  Farmer,et al.  Predicting chaotic time series. , 1987, Physical review letters.

[34]  A Hjelmfelt,et al.  Chemical implementation of neural networks and Turing machines. , 1991, Proceedings of the National Academy of Sciences of the United States of America.

[35]  A P Mills,et al.  Article for analog vector algebra computation. , 1999, Bio Systems.

[36]  Takashi Ikegami,et al.  Shapes and Self-Movement in Protocell Systems , 2009, Artificial Life.

[37]  Steen Rasmussen,et al.  Protocells : bridging nonliving and living matter , 2008 .

[38]  Christof Teuscher,et al.  COEL: A Cloud-Based Reaction Network Simulator , 2016, Front. Robot. AI.

[39]  Christof Teuscher,et al.  COEL: A Web-based Chemistry Simulation Framework , 2014, ArXiv.

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

[41]  D. Bray.  Protein molecules as computational elements in living cells , 1995, Nature.

[42]  Evgeny Katz,et al.  Biomolecular information processing : from logic systems to smart sensors and actuators , 2012 .

[43]  Tim J. Hutton,et al.  Evolvable Self-Reproducing Cells in a Two-Dimensional Artificial Chemistry , 2007, Artificial Life.

[44]  Christof Teuscher,et al.  Online Learning in a Chemical Perceptron , 2013, Artificial Life.

[45]  J. Ziegler,et al.  Artificial Chemistries-A Review , 2001 .

[46]  Matthew R. Lakin,et al.  Supervised Learning in an Adaptive DNA Strand Displacement Circuit , 2015, DNA.

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

[48]  Liqin Zhang,et al.  A survey of advancements in nucleic acid-based logic gates and computing for applications in biotechnology and biomedicine. , 2015, Chemical communications.

[49]  Christof Teuscher,et al.  Learning Two-Input Linear and Nonlinear Analog Functions with a Simple Chemical System , 2014, UCNC.

[50]  M. Win,et al.  Higher-Order Cellular Information Processing with Synthetic RNA Devices , 2008, Science.

[51]  Phil Husbands,et al.  Evolution of Associative Learning in Chemical Networks , 2012, PLoS Comput. Biol..

[52]  J. Ross,et al.  Chemical implementation and thermodynamics of collective neural networks. , 1992, Proceedings of the National Academy of Sciences of the United States of America.

[53]  Gérard Berry,et al.  The chemical abstract machine , 1989, POPL '90.

[54]  L. Glass,et al.  Oscillation and chaos in physiological control systems. , 1977, Science.

[55]  A. J. Lotka.  UNDAMPED OSCILLATIONS DERIVED FROM THE LAW OF MASS ACTION. , 1920 .

[56]  Amir F. Atiya,et al.  New results on recurrent network training: unifying the algorithms and accelerating convergence , 2000, IEEE Trans. Neural Networks Learn. Syst..

[57]  Shawn M. Douglas,et al.  A Logic-Gated Nanorobot for Targeted Transport of Molecular Payloads , 2012, Science.

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