Researchers Advance Toward Matching Patients with the Most Effective GLP-1 Drug for Personalized Obesity Treatment

New research, including a 2025 Mayo Clinic study, shows that machine-learning algorithms and genetic risk scores (such as Phenomix Sciences' MyPhenome test) can predict which patients are more likely to experience side effects (notably nausea) from GLP-1 therapies, enabling more personalized treatment decisions and improved drug development.1 11

Researchers are moving closer to developing benefit–risk scoring algorithms that support individualized drug selection, further advancing the field of precision obesity medicine.7

These predictive tools are designed to match the right patient with the right GLP-1 drug, improving tolerability, adherence, and patient outcomes while informing the design of clinical trials with better stratification by patient risk profiles.1 7

Large-scale real-world data demonstrate significant individual variability in both efficacy and side effects of GLP-1 drugs. Predictive models that integrate genetic, demographic, and clinical data are being developed to address this variability.5 9 10

The integration of AI and genetic data in clinical practice is beginning to shape treatment pathways for obesity and metabolic diseases, with several institutions (such as the Mayo Clinic and industry partners) validating these approaches in multi-center studies presented in 2025.1 11