A Machine-Learning Approach Identifies Rejuvenating Interventions in the Human Brain

Abstract

The increase in life expectancy has caused a rise in age-related brain disorders. Although brain rejuvenation is a promising strategy to counteract brain functional decline, systematic discovery methods for efficient interventions are lacking. A computational platform based on a transcriptional brain aging clock capable of detecting age- and neurodegeneration-related changes is developed. Applied to neurodegeneration-positive samples, it reveals that neurodegenerative disease presence and severity significantly increase predicted age. By screening 43840 transcriptional profiles of chemical and genetic perturbations, it identifies 453 unique rejuvenating interventions, several of which are known to extend lifespan in animal models. Additionally, the identified interventions include drugs already used to treat neurological disorders, Alzheimer's disease among them. A combination of compounds predicted by the platform reduced anxiety, improved memory, and rejuvenated the brain cortex transcriptome in aged mice. These results demonstrate the platform's ability to identify brain-rejuvenating interventions, offering potential treatments for neurodegenerative diseases.