Reservoir computing in chemical reaction networks

Institute for Molecules and Materials, Radboud University, Nijmegen, the Netherlands.

The cell is arguably the most amazing material that exist. Inspired by its unique functions – homeostasis, sensing, adaptation, growth – we are exploring routes to construct synthetic systems that capture some of the complexity of living systems.

Our focus is on complex networks of chemical reactions. In cells, they define how life works. We are familiar with the metabolic networks studied in biochemistry, and in recent decades many regularly recurring network motifs have been uncovered that are responsible for much of the functional behaviour in signalling or genetic networks. However, molecular ‘circuits’ are very delicate, and sensitive to changes in concentration, temperature, and so on. In this lecture, I will discuss strategies to ‘synthesize’ programmable reaction networks in microfluidic flow reactors, and we can use these networks as powerful reservoir computers.

Some relevant publications
[1]          S.N. Semenov, et al. Nature Chemistry, 2015, 7, 160-165  
[2]          A.S.Y. Wong, et al. J. Am. Chem. Soc. 2017, 139, 8146-8151
[3]          A.A.Pogodaev, et. al. J. Am. Chem. Soc. 2017, 139, 15296-15299.
[4]          te Brinke, et al. Nature Nanotech. 2018, 13, 849
[5]          Zheng, L.  et al. Nature Chemistry, 2019, 11, 359–366