IBM uses AI-powered model to predict age from gut microbiome samples
The team from IBM and University of California San Diego (UC San Diego) believes this sets the stage for future research on the microbiome’s role in accelerating or decelerating the aging process and in the susceptibility for age-related diseases.
“The accuracy of our results demonstrate the potential for applying machine learning and artificial intelligence techniques to better understand human microbiomes,” said co-author Ho-Cheol Kim, program director of the Artificial Intelligence for Healthy Living Program, a collaboration between IBM Research and UC San Diego under the IBM AI Horizons Network.
“Applying this technology to future microbiome studies could help unlock deeper insights into the correlation between how microbiomes influence our overall health and a wide range of diseases and disorders from neurological to cardiovascular and immune health.”
Considerable evidence suggests that the gut microbiome changes with age or even accelerates aging in adults.
Whether the age-related changes in the gut microbiome are more or less prominent than those for other body sites and whether predictions can be made about a person’s age from a microbiome sample remain unknown.
Study details
Headed up by co-senior authors Zhenjiang Zech Xu and Rob Knight, the team combined ten large studies from different countries to determine which body site’s microbiome could most accurately predict age.
Specifically, three different human microbiome populations on the skin, mouth and gut, were explored.
Mining data from these studies, the team gathered close to 9,000 microbiome samples from subjects aged between 18 and 90 years.
Next, machine learning was applied to predict age from the relative abundance of microbes within the sampled microbiomes.
Random forest regression models, tuned, trained and tested for the task, recaptured the previously association between the gut microbiome and age.
The researchers found the skin microbiome the most accurate indicator of chronological age, accurately calculating age within a range of 3.8 years.
Next was the oral microbiome with correct estimates of chronological age within 4.5 years. In contrast, the gut microbiome was the least accurate, correctly calculating age, on average within a range of 11.5 years.
‘Advancing future studies’
“This new ability to correlate microbes with age will help us advance future studies of the roles microbes play in the aging process and age-related diseases, and allow us to better test potential therapeutic interventions that target microbiomes,” says co-senior author Zhenjiang Zech Xu.
The team thinks that one potential reason the microbes living on our skin change so consistently as we age is due to the predictable changes in skin physiology that everyone experiences, such as decreased serum production and increased dryness.
“It was surprising to discover that the skin and oral microbiomes are much more predictive of age than gut microbiome,” write fellow team members Niina Haiminen, Anna-Paola Carrieri, and Ho-Cheol Kim in a blog.
“Promisingly, the hand and forehead skin microbiome age models generalised across cohorts and geographies, indicating studies from several sources could be combined in the future to potentially accelerate discovery from globally available microbiome data.”
The authors go on to discuss their findings that focus on the gut and oral microbes enriched in young subjects, which were found to be more abundant and more prevalent than microbes enriched in the old subjects.
“This suggests a model where aging occurs in tandem with the loss of key microbes over a lifetime,” the authors write.
“This observation sets the stage for future research on the role of the microbiome in the aging process. Taken together, the results demonstrate that accurate and generalisable indicators of age can be derived from using machine learning on microbiome data.”
Source: mSystems
Published online ahead of print: DOI: 10.1128/mSystems.00630-19
“Human Skin, Oral, and Gut Microbiomes Predict Chronological Age.”
Authors: Shi Huang et al.
