Computing within bacteria: Programming of bacterial behavior by means of a plasmid encoding a perceptron neural network

Nonlinear classifiers for wet-neuromorphic computing using gene regulatory neural network

The gene regulatory network (GRN) of biological cells governs a number of key functionalities that enable them to adapt and survive through different environmental conditions. Close observation of the GRN shows that the structure and operational principles resemble an artificial neural network (ANN), which can pave the way for the development of wet-neuromorphic computing systems. Genes are integrated into gene-perceptrons with transcription factors (TFs) as input, where the TF concentration relative to half-maximal RNA concentration and gene product copy number influences transcription and translation via weighted multiplication before undergoing a nonlinear activation function. This process yields protein concentration as the output, effectively turning the entire GRN into a gene regulatory neural network (GRNN). In this paper, we establish nonlinear classifiers for molecular machine learning using the inherent sigmoidal nonlinear behavior of gene expression. The eigenvalue-based stability analysis, tailored to system parameters, confirms maximum-stable concentration levels, minimizing concentration fluctuations and computational errors. Given the significance of the stabilization phase in GRNN computing and the dynamic nature of the GRN, alongside potential changes in system parameters, we utilize the Lyapunov stability theorem for temporal stability analysis. Based on this GRN-to-GRNN mapping and stability analysis, three classifiers are developed utilizing two generic multilayer sub-GRNNs and a sub-GRNN extracted from the Escherichia coli GRN. Our findings also reveal the adaptability of different sub-GRNNs to suit different application requirements.

Synthetic biological neural networks: From current implementations to future perspectives

Artificial neural networks, inspired by the biological networks of the human brain, have become game-changing computing models in modern computer science. Inspired by their wide scope of applications, synthetic biology strives to create their biological counterparts, which we denote synthetic biological neural networks (SYNBIONNs). Their use in the fields of medicine, biosensors, biotechnology, and many more shows great potential and presents exciting possibilities. So far, many different synthetic biological networks have been successfully constructed, however, SYNBIONN implementations have been sparse. The latter are mostly based on neural networks pretrained in silico and being heavily dependent on extensive human input. In this paper, we review current implementations and models of SYNBIONNs. We briefly present the biological platforms that show potential for designing and constructing perceptrons and/or multilayer SYNBIONNs. We explore their future possibilities along with the challenges that must be overcome to successfully implement a scalable in vivo biological neural network capable of online learning.

Revealing gene regulation-based neural network computing in bacteria

Bacteria are known to interpret a range of external molecular signals that are crucial for sensing environmental conditions and adapting their behaviors accordingly. These external signals are processed through a multitude of signaling transduction networks that include the gene regulatory network (GRN). From close observation, the GRN resembles and exhibits structural and functional properties that are similar to artificial neural networks. An in-depth analysis of gene expression dynamics further provides a new viewpoint of characterizing the inherited computing properties underlying the GRN of bacteria despite being non-neuronal organisms. In this study, we introduce a model to quantify the gene-to-gene interaction dynamics that can be embedded in the GRN as weights, converting a GRN to gene regulatory neural network (GRNN). Focusing on Pseudomonas aeruginosa, we extracted the GRNN associated with a well-known virulence factor, pyocyanin production, using an introduced weight extraction technique based on transcriptomic data and proving its computing accuracy using wet-lab experimental data. As part of our analysis, we evaluated the structural changes in the GRNN based on mutagenesis to determine its varying computing behavior. Furthermore, we model the ecosystem-wide cell-cell communications to analyze its impact on computing based on environmental as well as population signals, where we determine the impact on the computing reliability. Subsequently, we establish that the individual GRNNs can be clustered to collectively form computing units with similar behaviors to single-layer perceptrons with varying sigmoidal activation functions spatio-temporally within an ecosystem. We believe that this will lay the groundwork toward molecular machine learning systems that can see artificial intelligence move toward non-silicon devices, or living artificial intelligence, as well as giving us new insights into bacterial natural computing.

Design of Intelligent Neuro-Supervised Networks for Brain Electrical Activity Rhythms of Parkinson's Disease Model

The objective of this paper is to present a novel design of intelligent neuro-supervised networks (INSNs) in order to study the dynamics of a mathematical model for Parkinson's disease illness (PDI), governed with three differential classes to represent the rhythms of brain electrical activity measurements at different locations in the cerebral cortex. The proposed INSNs are constructed by exploiting the knacks of multilayer structure neural networks back-propagated with the Levenberg-Marquardt (LM) and Bayesian regularization (BR) optimization approaches. The reference data for the grids of input and the target samples of INSNs were formulated with a reliable numerical solver via the Adams method for sundry scenarios of PDI models by way of variation of sensor locations in order to measure the impact of the rhythms of brain electrical activity. The designed INSNs for both backpropagation procedures were implemented on created datasets segmented arbitrarily into training, testing, and validation samples by optimization of mean squared error based fitness function. Comparison of outcomes on the basis of exhaustive simulations of proposed INSNs via both LM and BR methodologies was conducted with reference solutions of PDI models by means of learning curves on MSE, adaptive control parameters of algorithms, absolute error, histogram error plots, and regression index. The outcomes endorse the efficacy of both INSNs solvers for different scenarios in PDI models, but the accuracy of the BR-based method is relatively superior, albeit at the cost of slightly more computations.

Realizing Molecular Machine Learning through Communications for Biological AI: Future Directions and Challenges

Artificial Intelligence (AI) and Machine Learning (ML) are weaving their way into the fabric of society, where they are playing a crucial role in numerous facets of our lives. As we witness the increased deployment of AI and ML in various types of devices, we benefit from their use into energy-efficient algorithms for low powered devices. In this paper, we investigate a scale and medium that is far smaller than conventional devices as we move towards molecular systems that can be utilized to perform machine learning functions, i.e., Molecular Machine Learning (MML). Fundamental to the operation of MML is the transport, processing, and interpretation of information propagated by molecules through chemical reactions. We begin by reviewing the current approaches that have been developed for MML, before we move towards potential new directions that rely on gene regulatory networks inside biological organisms as well as their population interactions to create neural networks. We then investigate mechanisms for training machine learning structures in biological cells based on calcium signaling and demonstrate their application to build an Analog to Digital Converter (ADC). Lastly, we look at potential future directions as well as challenges that this area could solve.

A computational design of a programmable biological processor

Basic synthetic information processing structures, such as logic gates, oscillators and flip-flops, have already been implemented in living organisms. Current implementations of these structures have yet to be extended to more complex processing structures that would constitute a biological computer. We make a step forward towards the construction of a biological computer. We describe a model-based computational design of a biological processor that uses transcription and translation resources of the host cell to perform its operations. The proposed processor is composed of an instruction memory containing a biological program, a program counter that is used to address this memory, and a biological oscillator that triggers the execution of the next instruction in the memory. We additionally describe the implementation of a biological compiler that compiles a sequence of human-readable instructions into ordinary differential equation-based models, which can be used to simulate and analyse the dynamics of the processor. The proposed implementation presents the first programmable biological processor that exploits cellular resources to execute the specified instructions. We demonstrate the application of the described processor on a set of simple yet scalable biological programs. Biological descriptions of these programs can be produced manually or automatically using the provided compiler.