Predicting antimicrobial resistance in Pseudomonas aeruginosa with machine learning-enabled molecular diagnostics
By combining genomic and transcriptomic data with machine learning, researchers achieved high accuracy in predicting antibiotic resistance in Pseudomonas aeruginosa for several antibiotics. Gene expression data significantly improved prediction, especially for ceftazidime, meropenem, and tobramycin, and classifiers performed better when MIC values were further from breakpoints.