Dr. Qingsong Zhu, the COO of Insilico Medicine, discussed the use of deep learning in creating biomarkers for aging. Initially discussing existing clocks and the problems with animal translation, he went on to discuss what sorts of markers are ideal for age-related research and the details of training and testing a model that works with these markers, showing that a deep model compares favorably to other models.

He also used his model to show that smoking does, in fact, cause accelerated aging.