Genetic circuits combined with machine learning provides fast responding living sensors

Behide Saltepe; Eray Ulaş Bozkurt; Murat Alp Güngen; A Ercüment Çiçek; Urartu Özgür Şafak Şeker

doi:10.1016/j.bios.2021.113028

Genetic circuits combined with machine learning provides fast responding living sensors

Biosens Bioelectron. 2021 Apr 15:178:113028. doi: 10.1016/j.bios.2021.113028. Epub 2021 Jan 23.

Authors

Behide Saltepe¹, Eray Ulaş Bozkurt¹, Murat Alp Güngen¹, A Ercüment Çiçek², Urartu Özgür Şafak Şeker³

Affiliations

¹ UNAM - Institute of Materials Science and Nanotechnology, National Nanotechnology Research Center, Bilkent University, 06800, Bilkent-Ankara, Turkey.
² Computer Engineering Department, Bilkent University, 06800, Bilkent-Ankara, Turkey; Computer Biology Department, Carnegie Mellon University, 15213, Pittsburgh, PA, USA. Electronic address: cicek@cs.bilkent.edu.tr.
³ UNAM - Institute of Materials Science and Nanotechnology, National Nanotechnology Research Center, Bilkent University, 06800, Bilkent-Ankara, Turkey. Electronic address: urartu@bilkent.edu.tr.

PMID: 33508538
DOI: 10.1016/j.bios.2021.113028

Abstract

Whole cell biosensors (WCBs) have become prominent in many fields from environmental analysis to biomedical diagnostics thanks to advanced genetic circuit design principles. Despite increasing demand on cost effective and easy-to-use assessment methods, a considerable amount of WCBs retains certain drawbacks such as long response time, low precision and accuracy. Here, we utilized a neural network-based architecture to improve the features of WCBs and engineered a gold sensing WCB which has a long response time (18 h). Two Long-Short Term-Memory (LSTM)-based networks were integrated to assess both ON/OFF and concentration dependent states of the sensor output, respectively. We demonstrated that binary (ON/OFF) network was able to distinguish between ON/OFF states as early as 30 min with 78% accuracy and over 98% in 3 h. Furthermore, when analyzed in analog manner, we demonstrated that network can classify the raw fluorescence data into pre-defined analyte concentration groups with high precision (82%) in 3 h. This approach can be applied to a wide range of WCBs and improve rapidness, simplicity and accuracy which are the main challenges in synthetic biology enabled biosensing.

Keywords: Living sensors; Machine learning; Neural networks; Synthetic biology; Whole-cell biosensors.

MeSH terms

Biosensing Techniques*
Gene Regulatory Networks
Machine Learning
Neural Networks, Computer
Synthetic Biology