Deep-learning models are ubiquitous in our everyday lives, examples ranging from applications in computer vision, natural language processing, recommender systems, and the development of the self-driving car. It therefore does not come as a surprise that deep learning also impacts scientific discovery. In fact, among the various fields that have particularly benefited from such models are the scientific disciplines related to drug discovery. Activity prediction, ADMET modeling, target identification, and de-novo drug design are a few of the areas where these modern machine-learning methods have shown promise.
While deep learning approaches in many cases outperform other approaches that are commonly used in the field, their predictions are commonly considered as obscure or hard to interpret. Explainable Artificial Intelligence (XAI) methods aim to overcome this limitation.
Owing to the importance and timeliness of this subject, Dr. José Jimenez joined the RETHINK initiative at ETH Zurich as a postdoctoral researcher, in collaboration with Boehringer Ingelheim Pharma. Dr. Jimenez has a background in statistics and operations research, and obtained his Ph.D. on the subject of machine-learning applications to drug discovery. His main research focus at RETHINK will be on the development and exploration of XAI models in the context of drug design.