Podcast about our Human Phenotype Project humanphenotypeproject.org/ describing our large-scale cohort, the clinical and genomic data collected, the goals, and the AI tools we're developing to model future health trajectories of people Listen here: tinyurl.com/hme5bsz8 What's even more amazing, is that the podcast was generated in minutes using Google's new Notebook ML tool @JeffDean based only on a written pdf of our project. The podcast is coherent, fluent, and interesting. Mind blowing

Our new @NatureAging paper: We built 14 "biological clocks" representing the biological age and rate of aging of 14 major body systems using data from 13,000 people of the Human Phenotype Project Key findings: 1. Biological clocks are clinically meaningful: People with higher biological age for some body system are at greater risk for disease related to that body system (e.g., higher aging rate of insulin resistance associates with higher risk for diabetes) 2. Surprisingly, biological clocks for different body systems are mostly independent, i.e., a person with a higher rate of aging for one body system does not necessarily have higher aging rates in other body systems 3. Males and females age differently: While males age linearly, females exhibit a sharp non-linear increase between the ages 50-55. This sharp increase is associated with menopause. When comparing pre- and post-menopausal females of the same chronological age, we find that post-menopausal females have higher biological age for several body systems Paper: tinyurl.com/5esb3f5h Data: humanphenotypeproject.org Access our aging models: tinyurl.com/42yctten Great job by @Lee_Reicher1, @NoamBar7, @nastya_godneva, Yotam Reisner, @ZahaviLiron, Nir Shahaf, @AdinaBlueBird, in collaboration with Raja Dhir