Use of metabolomics and machine learning to counter antibiotic resistance
Émanuel Paré, Francis Brière, Thibaud Godon, Maxime Déraspe, Émilie Pic, Baptiste Bauvin and Jacques Corbeil
Université Laval, departments of computing science and molecular medicine, INAF and NUTRISS
Antibiotic resistance is a growing global issue that urgently requires mobilization. Artificial intelligence can be used to identify innovative solutions to the antibiotic resistance. Our objective was to develop rapid and precise tests to evaluate antibiotic resistance, using mass spectrometry coupled with machine learning (MAGITICS project from the JPIAMR). Additionally, we aim to facilitate the development of new antibiotics using artificial intelligence approaches to analyze antimicrobial peptide databases and identify candidates with clinical development potential. We intend to use the generative flow net (1) and the robust set covering machine to identify such candidates by modulating the exploration/utilization ratio for new drug discovery.