Chandrasekhar Hall

Building computers and artificial neural networks with engineered bacteria that can compute, solve mazes and add and subtract numbers

Sangram Bagh

Biophysics and Structural Genomics Division, Saha Institute of Nuclear Physics, Kolkata

Performing cellular computations with engineered bacteria has enormous importance in biocomputer technology development at the micron scale, where microprocessor-based computers have limitations due to energy, cost and technological constraints. Here, we designed and built artificial neural networks and biocomputers with molecular engineered E.coli. In this work, we have adapted the basic concept of artificial neural networks (ANNs) and designed, built and optimized synthetic genetic circuits in E.coli, where engineered bacteria worked as artificial `neuro-synapses'. The engineered bacteria in a culture work as an artificial neural network. Here, the input chemicals are processed by engineered gene regulatory networks distributed among multiple artificial bacterial neurons, and appropriate output is manifested with an expression pattern of multiple fluorescent proteins. The bacterial neurons are connected by chemical wires. First, we established a set of rules on how a computing function can be mapped onto a bacteria-based ANN and how those mathematical forms can be translated into molecular design principle. Applying those rules we experimentally demonstrated a broadly applicable single-layer ANN type architecture with molecular-engineered E.coli to perform complex irreversible computing like multiplexing, de-multiplexing, encoding, decoding, majority functions, and reversible computing like Feynman, double Feynman and Fredkin gates. Further, we expanded the capability of bacterial ANN and built a configurable cellular system, that can add and subtract binary numbers. Adapting the idea of distributed computing and applying it to cellular computing, we build a biocomputer, that can solve simple maze problems. To our knowledge, this is the first artificial neural network (ANN) with genetically engineered cells. This work may have significance in establishing a new platform for cellular computing and in transforming bacterial cells into ANN-enabled hardware for various applications.