Premium
A high‐throughput and machine learning resistance monitoring system to determine the point of resistance for Escherichia coli with tetracycline: Combining UV‐visible spectrophotometry with principal component analysis
Author(s) -
Chapman James,
OrrellTrigg Rebecca,
Kwoon Ki Y.,
Truong Vi K.,
Cozzolino Daniel
Publication year - 2021
Publication title -
biotechnology and bioengineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.136
H-Index - 189
eISSN - 1097-0290
pISSN - 0006-3592
DOI - 10.1002/bit.27664
Subject(s) - principal component analysis , escherichia coli , tetracycline , spectrophotometry , chemometrics , ultraviolet visible spectroscopy , linear discriminant analysis , biological system , optical density , partial least squares regression , analytical chemistry (journal) , chromatography , artificial intelligence , chemistry , computer science , optics , microbiology and biotechnology , antibiotics , biology , physics , machine learning , biochemistry , organic chemistry , gene
Abstract UV‐visible spectroscopy (UV‐Vis) is routinely used in microbiology as a tool to check the optical density (OD) pertaining to the growth stages of microbial cultures at the single wavelength of 600 nm, better known as the OD 600 . Typically, modern UV‐Vis spectrophotometers can scan in the region of approximately 200–1000 nm in the electromagnetic spectrum, where users do not extend the use of the instrument's full capability in a laboratory. In this study, the full potential of UV‐Vis spectrophotometry (multiwavelength collection) was used to examine bacterial growth phases when treated with antibiotics showcasing the ability to understand the point of resistance when an antibiotic is introduced into the media and therefore understand the biochemical changes of the infectious pathogens. A multiplate reader demonstrated a high throughput experiment (96 samples) to understand the growth of Escherichia coli when varied concentrations of the antibiotic tetracycline was added into the well plates. Principal component analysis (PCA) and partial least squares discriminant analysis were then used as the data mining techniques to interpret the UV‐Vis spectral data and generate machine learning “proof of principle” for the UV‐Vis spectrophotometer plate reader. Results from this study showed that the PCA analysis provides an accurate yet simple visual classification and the recognition of E. coli samples belonging to each treatment. These data show significant advantages when compared to the traditional OD 600 method where we can now understand biochemical changes in the system rather than a mere optical density measurement. Due to the unique experimental setup and procedure that involves indirect use of antibiotics, the same test could be used for obtaining practical information on the type, resistance, and dose of antibiotic necessary to establish the optimum diagnosis, treatment, and decontamination strategies for pathogenic and antibiotic resistant species.