A team of researchers based in Qatar is developing an artificial intelligence (AI) system designed to predict patient responses to antibiotics before treatment begins, potentially reducing harmful side effects and improving therapeutic outcomes. The project, expected to run for three years, is co-led by Dr. Jithesh Puthenveetil, a professor at Hamad Bin Khalifa University’s (HBKU) College of Health and Life Sciences and associate dean for Education and Student Affairs.
The initiative involves a collaboration between HBKU, Hamad Medical Corporation (HMC), and the Qatar Precision Health Institute (QPHI), with funding provided by the Qatar Research, Development and Innovation Council. By integrating AI, genomics, and precision medicine, the research aims to tailor antibiotic treatments to individual patients based on their genetic profiles.
“Two patients may receive the same medicine, but one benefits while the other may experience little improvement or even serious side effects,” Puthenveetil explained. He emphasized that genetic variation often underlies such differences in drug response, a concept central to the field of pharmacogenomics. This understanding allows for the customization of medical treatments to fit the specific genetic makeup of each patient.
The project builds on earlier work from the Qatar Genome Programme Research Consortium, in which Puthenveetil previously led pharmacogenomics research. The current phase adds a deeper layer by combining genomic data with patients’ clinical records and AI-driven analysis to improve the accuracy of predictions about drug response.
Rather than addressing all infectious diseases, the study focuses specifically on antibiotics commonly prescribed in Qatari hospitals for infections such as pneumonia, sepsis, and skin and soft tissue infections. Approximately 600 patients will be recruited for the study. Researchers will collect samples—including blood, urine, and stool—from these individuals for genome sequencing, metabolomic analysis, and microbiome profiling. These biological data will be cross-referenced with electronic medical records documenting medication histories, treatment outcomes, and any adverse reactions.
Puthenveetil highlighted that the integration of large-scale biological and clinical data sets represents the project’s key innovation. “Artificial intelligence allows us to integrate these different data layers and identify patterns that would otherwise remain hidden,” he said. By uncovering these patterns, the research team hopes to enable healthcare providers to make more informed decisions about antibiotic use, potentially reducing trial-and-error prescribing and minimizing risks associated with inappropriate medication.