Machine Learning Can Help Predict Healthy Aging

A study published in Genome Biology from researchers at the Salk Institute used machine learning to identify key molecular signatures across all age groups that predict aging, reports Earth.com . "We want to develop algorithms that can predict healthy aging and non-healthy aging, and try to find the differences," said Salk's Saket Navlakha. The team analyzed dermal fibroblast skin cells from 133 individuals aged one year to 94 years with about 13 people per decade of age to compile a large enough representation of the different age groups. RNA sequencing was used to identify any biomarkers of aging and the genes that were turned on or off within the cells, and then custom machine learning algorithms analyzed the sequencing data. The algorithms could predict a person's age from their cells within a range of eight years. Tests on subjects with progeria showed the cells were at least 10 years older than the actual age of the subjects they came from. "The fact that our system can predict this kind of aging shows that this model is starting to get at the true underpinnings of biological age," said Salk's Jason Fleischer.