Abstract

Building AI Models of Patient-specific Drug Side Effect Predictions

Planning for unforeseen side effects of medication when choosing management options for diseases like cancer is a perennial problem for clinicians. This problem is exacerbated for designer diseases in which each patient can present completely differently. Currently, the only information about side effects we have is from the original clinical trials on the medication in animal and human models. However, automated models, specifically utilizing the benefits of machine learning, can help advance precision medicine by predicting side effects in silico. Recent advances in applying machine learning techniques to chemical structures of molecules enables this. In this study, we provide an early proof-of-concept implementation that transforms molecular structures to a list of potential predicted side effects. We design an extended architecture that pairs molecular structures with patient profiles to generate patient-specific recommendations.Clinical relevance: The use of AI tools to predict side effects for drugs can become a powerful tool for doctors over the coming decade. By combining molecular information about drugs with patient specific knowledge, AI methods can predict patient-specific side effects with high accuracy.