image: Insilico Medicine, Inc., a company applying latest advances in deep learning to biomarker development, drug discovery and aging research, launched Aging.AI 2.0. Comparing Aging.AI 1.0 using 41 blood biochemistry biomarkers, Aging.AI 2.0 uses just 33 parameters from the blood test and has slightly higher mean absolute error. view more
Credit: Insilico Medicine
Monday, November 14, 2016, Baltimore, MD - Today, Insilico Medicine, Inc., a company applying latest advances in deep learning to biomarker development, drug discovery and aging research, launched Aging.AI 2.0, the blood biochemistry predictor of human age. Capitalizing on the success of Aging.AI 1.0 platform, using just 41 blood biochemistry biomarkers launched in January 2016 and tested by thousands of people, the Aging.AI 2.0 allows users to use just 33 parameters from their recent blood test to guess their chronological age. The system is available for beta testing via http://www.Aging.AI .
The Aging.AI 2.0 has slightly higher mean absolute error than the previous version; however, it covers more population groups and works slightly better on the long tail of the older population. The research study behind the Aging.AI system was published in a leading peer-reviewed journal in the field of aging: Putin, et al, "Deep biomarkers of human aging: Application of deep neural networks to biomarker development." Aging 8, no. 5 (2016): 1-021 and recent studies demonstrated that these markers are population-specific. At the recent "3rd International Aging Research for Drug Discovery" conference in Basel, Switzerland, Dr. Mun Yew Wong, the CEO of Asia Genomics presented the first insights into a study demonstrating that certain population groups in Asia are guessed younger in older age by the deep neural networks trained on Eastern European population and the mean absolute error is higher.
"Deep Learning with no doubt has a huge potential in healthcare, but unfortunately, very few groups are applying it to aging research. Aging is one of the most complex and multifactorial processes killing millions every year and causing more suffering than any other known disease. We are developing deep integrated biomarkers of aging that incorporate blood biochemistry, transcriptomics and even imaging data to be able to track the effectiveness of the various interventions we are developing", said Alex Zhavoronkov, PhD, CEO of Insilico Medicine.
Insilico Medicine also sees the applications of deep learned biomarkers of aging in multiple applications including clinical trials enrollment, clinical practice and regular health checkups.
"Old-school physicians were trained to guess the age of the patient the moment he or she walked into the office and if the patient looked significantly older than the chronological age, more extensive testing was advised. This may be the case with Aging.AI, since what we are really looking for when training the DNNs to guess the age of reasonably healthy people is the biomarker of health status. And even though the system may guess your age with significant error when you first use it, what we want to study is the differential changes for each individual patient so people could monitor their health and adjust their lifestyle", said Alex Aliper, president of European operations at Insilico Medicine.
Aging.AI 2.0 was trained on more samples from North America, and Central Europe and may demonstrate lower error rates on across population groups than Aging.AI 1.0. Insilico Medicine is constantly looking for collaborators with large data sets to develop better biomarkers of aging and disease. Please contact Insilico Medicine for collaboration opportunities.
"Our research team is primarily focused on developing transcriptomic blood-derived and tissue-specific deep learned biomarkers of aging and disease trained on a large number of gene expression datasets and blood biochemistry data is not our primary focus. However, encouraged by the success of the first version of Aging.AI, we decided to improve our algorithm and change the interface of the web-version to made it more user-friendly. The system is easy to use I hope that this would stimulate more people to participate in aging research and to pay more attention to their own health. We are working on integrating the blood biochemistry data with gene expression data in order to build a comprehensive, biologically-relevant biomarker of aging" - said Polina Mamoshina, a Research Scientist of Pharmaceutical Artificial Intelligence division of Insilico Medicine.
###
About Insilico Medicine
Insilico Medicine, Inc. is a bioinformatics company located at the Emerging Technology Centers at the Johns Hopkins University Eastern campus in Baltimore with R&D resources in Belgium, Russia and Poland hiring talent through hackathons and competitions. It utilizes advances in genomics, big data analysis and deep learning for in silico drug discovery and drug repurposing for aging and age-related diseases. The company pursues internal drug discovery programs in cancer, Parkinson's, Alzheimer's, sarcopenia and geroprotector discovery. Through its Pharma.AI division the company provides advanced machine learning services to biotechnology, pharmaceutical and skin care companies. Brief company video: https://www.youtube.com/watch?v=l62jlwgL3v8