Applications of Deep Learning Methods for Aging Research & Other Areas of Healthcare
While there is still skepticism with regard to applications of deep learning to biomarker development and drug discovery using blood biochemistry and transcriptomic data, there are multiple applications that show promise. Recently we used an ensemble of feed forward deep neural networks to build a predictor of chronological age and gender trained on a data set of 920,000 common blood biochemistry and clinical cell count data samples of 41 parameters each achieving F1 scores of roughly 0.84 with epsilon of 10 years and F1 of 0.96 for gender. The predictor is available at www.Aging.AI.
Alex Zhavoronkov, PhD, is the CEO of Insilico Medicine, CSO of the Biogerontology Research Foundation, director of the International Aging Research Portfolio (IARP), head of the regenerative medicine lab at the Center for Pediatric Hematology, Oncology and Immunology in Moscow and adjunct professor of the Moscow Institute of Physics and Technology. Previously he served as the director of ATI Technologies and CTO of NeuroG Neuroinformatics. He is the author of "The Ageless Generation: How advances in biomedicine will transform the global economy" (Palgrave Macmillan, 2013) and co-organizer of the annual Aging Research Forum at the Basel Life Science Week.