Application of Deep Neural Networks for Human Ageing Biomarker Investigation
The ensemble approach to predict chronological age has been previously explored in East European populations. The best performing deep neural net demonstrated 81.5% accuracy with a MAE of 6.07 years. We set out to explore if there are common predictors of age in Asian blood biochemistry markers. We have collected 30,000 blood biochemistry markers from age range of 2 years to 99 years. Using a previously trained Deep Neural Net, we observed a MAE of 10.74 between predicted and chronological age. The initial findings suggest that deep biomarkers of ageing are population specific.
Dr. WONG Mun Yew has experience in clinical medicine, investments, operations and being an entrepreneur. He practised as a medical doctor, before joining Bio*One Capital, a healthcare investment fund, before becoming Senior Vice President at EDB Investments Pte Ltd where he led EDBI’s cleantech investment team. After leaving EDBI, he joined US Silicon Valley venture capital firm Formation8 Partners as Operating Partner and led its activities in Singapore and South-East Asia from 2013 to 2016, while founding Asia Genomics in 2014 as an entrepreneur. He has served as a director for Codexis Inc., Amaranth Medical Inc., Silecs International Pte Ltd, and was board observer in Fluidigm Corp, KaloBios Pharmaceuticals, Revance Therapeutics and Innovalight Inc. He has also gained operations experience in healthcare services as Vice President (International Operations) at Parkway Pantai Limited, and spent time in their China operations.