Online Learning in a Chemical Perceptron

Banda, Peter; Teuscher, Christof; Lakin, Matthew R.

doi:10.1162/artl_a_00105

Cited by 30 publications

(40 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Banda et al [48,49] have simulated several kinds of trainable perceptron using artifical chemistry, which is conceptually similar to our work but was designed using more abstract chemical reactions. It has been shown that such arbitrary systems of chemical reactions can be faithfully implemented in a biochemical setting by compiling them into an enlarged system of DNA strand displacement reactions [50].…”

Section: Related Workmentioning

confidence: 98%

Design of a biochemical circuit motif for learning linear functions

Lakin

Minnich

Lane

et al. 2014

J. R. Soc. Interface.

Self Cite

View full text Add to dashboard Cite

Learning and adaptive behaviour are fundamental biological processes. A key goal in the field of bioengineering is to develop biochemical circuit architectures with the ability to adapt to dynamic chemical environments. Here, we present a novel design for a biomolecular circuit capable of supervised learning of linear functions, using a model based on chemical reactions catalysed by DNAzymes. To achieve this, we propose a novel mechanism of maintaining and modifying internal state in biochemical systems, thereby advancing the state of the art in biomolecular circuit architecture. We use simulations to demonstrate that the circuit is capable of learning behaviour and assess its asymptotic learning performance, scalability and robustness to noise. Such circuits show great potential for building autonomous in vivo nanomedical devices. While such a biochemical system can tell us a great deal about the fundamentals of learning in living systems and may have broad applications in biomedicine (e.g. autonomous and adaptive drugs), it also offers some intriguing challenges and surprising behaviours from a machine learning perspective.

show abstract

Section: Related Workmentioning

confidence: 98%

Design of a biochemical circuit motif for learning linear functions

Lakin

Minnich

Lane

et al. 2014

J. R. Soc. Interface.

Self Cite

View full text Add to dashboard Cite

show abstract

“…To describe and highlight the main features of the ASP, we will compare it with the WRP [4] in the remainder of this paper. The WRP has two types of species and reaction symmetries.…”

Section: Model and Designmentioning

confidence: 99%

“…P reversibility, to prevent an infinite loop, we must precisely time the recovery phase and have a special species catalysing (guarding) the comparison. We implemented this rather cumbersome mechanism for the WLP as described previously [4]. An improved symmetric strategy is to compare the concentrations of species indirectly by their impact on the concurrent reactions they catalyse, and so annihilation occurs at the level of complementary products.…”

Section: (I) the Problem Of Negative Numbersmentioning

confidence: 99%

See 1 more Smart Citation

Training an asymmetric signal perceptron through reinforcement in an artificial chemistry

Banda

Teuscher

Stefanović

2014

J. R. Soc. Interface.

Self Cite

View full text Add to dashboard Cite

ResearchCite this article: Banda P, Teuscher C, Stefanovic D. 2014 Training an asymmetric signal perceptron through reinforcement in an artificial chemistry. J. R. Soc. State-of-the-art biochemical systems for medical applications and chemical computing are application-specific and cannot be reprogrammed or trained once fabricated. The implementation of adaptive biochemical systems that would offer flexibility through programmability and autonomous adaptation faces major challenges because of the large number of required chemical species as well as the timing-sensitive feedback loops required for learning. In this paper, we begin addressing these challenges with a novel chemical perceptron that can solve all 14 linearly separable logic functions. The system performs asymmetric chemical arithmetic, learns through reinforcement and supports both Michaelis-Menten as well as mass-action kinetics. To enable cascading of the chemical perceptrons, we introduce thresholds that amplify the outputs. The simplicity of our model makes an actual wet implementation, in particular by DNA-strand displacement, possible.

show abstract

“…This has allowed us to study the same system and modify its simulation dynamics from separate campuses. We have successfully applied COEL as a sole tool to model and evaluate various chemical perceptrons (Banda et al, 2013, chemical delay lines, and time-series learners Moles et al, 2015).…”

Section: Introduction and Related Workmentioning

confidence: 99%

COEL: A Cloud-Based Reaction Network Simulator

Banda

Teuscher

2016

Front. Robot. AI

Self Cite

View full text Add to dashboard Cite

Chemical Reaction Networks (CRNs) are a formalism to describe the macroscopic behavior of chemical systems. We introduce COEL, a web-and cloud-based CRN simulation framework, which does not require a local installation, runs simulations on a large computational grid, provides reliable database storage, and offers a visually pleasing and intuitive user interface. We present an overview of the underlying software, the technologies, and the main architectural approaches employed. Some of COEL's key features include ODE-based simulations of CRNs and multicompartment reaction networks with rich interaction options, a built-in plotting engine, automatic DNA-strand displacement transformation and visualization, SBML/Octave/Matlab export, and a built-in genetic-algorithm-based optimization toolbox for rate constants. COEL is an open-source project hosted on GitHub (doi:10.5281/zenodo.46544), which allows interested research groups to deploy it on their own sever. Regular users can simply use the web instance at no cost at coel-sim.org. The framework is ideally suited for a collaborative use in both research and education.

show abstract

Online Learning in a Chemical Perceptron

Cited by 30 publications

References 42 publications

Design of a biochemical circuit motif for learning linear functions

Design of a biochemical circuit motif for learning linear functions

Training an asymmetric signal perceptron through reinforcement in an artificial chemistry

COEL: A Cloud-Based Reaction Network Simulator

Contact Info

Product

Resources

About