Development of Nanomaterials and Machine Learning for Antimicrobial Resistance Spectroscopy
Dr.
Malama
Chisanga
Postdoctoral
Fellow
Department
of
Chemistry
Universite
de
Montreal
Monday,
January
23,
2023
10:30
a.m.
In-person: C2- 361 (Reading Room) and Online via MS Teams
Abstract: Vibrational spectroscopy and nanoscience combined with advanced chemometric tools exhibit unique capabilities to characterise chemical composition of complex samples. Unlike standard platforms such as mass spectrometry and fluorescence, vibrational spectroscopy is non-destructive, rapid and allows multiplex analysis of target analytes in their natural context in situ with little or no sample preparation. In this seminar, we will show how we developed nanoscience, analytical assays and machine learning to allow for surface-enhanced Raman scattering (SERS)-based detection of analytes of industrial and medical relevance. SERS has been designed for direct measurements and quantification of molecular fingerprints of diagnostic metabolites, and when combined with isotope labelling, identifies bacteria associated with the degradation of chemical pollutants. We also investigate the use of 2D materials, functionalised nanoparticles and microfluidics for simultaneous multiplexed detection of COVID-19-specific antibodies in 30 minutes with >95% sensitivity and specificity in clinical samples. This was then extended to the phenotypic screening of molecular changes (lipids, amino acids, lipids, proteins, etc.) linked to SARS-CoV-2 infection and recovery in longitudinal serum, and how spectral data and machine learning were applied to classify PCR-positive COVID-19 patients and negative control subjects. Next, we discuss the future research plans centred on the development of high-performance nanoparticles, surface chemistry, experimental protocols and machine learning tools to design new SERS and Raman imaging strategies for early detection of infectious diseases and antimicrobial resistance (AMR), the silent pandemic of the 21st century. We aim to probe AMR in volatile molecules, and in biofilms formed on medical devices that account for >80% of hospital-acquired infections (e.g., urinary tract infections, sepsis), with potential for point of use diagnosis. Qualitative and quantitative analysis of antibiotics in environmental samples will be discussed, and how this could be used to assess the prevalence of AMR. We hope to apply spectral molecular profiles of infection and drug resistance information for rapid identification of resistant infectious diseases to guide timely prescription of suitable and personalised antibiotics in order to improve patient outcomes and minimise the spread of AMR.