Age and life expectancy clocks based on machine learning analysis of mouse frailty

Abstract

The identification of genes and interventions that slow or reverse aging is hampered by the lack of non-invasive metrics that can predict the life expectancy of pre-clinical models. Frailty Indices (FIs) in mice are composite measures of health that are cost-effective and non-invasive, but whether they can accurately predict health and lifespan is not known. Here, mouse FIs are scored longitudinally until death and machine learning is employed to develop two clocks. A random forest regression is trained on FI components for chronological age to generate the FRIGHT (Frailty Inferred Geriatric Health Timeline) clock, a strong predictor of chronological age. A second model is trained on remaining lifespan to generate the AFRAID (Analysis of Frailty and Death) clock, which accurately predicts life expectancy and the efficacy of a lifespan-extending intervention up to a year in advance. Adoption of these clocks should accelerate the identification of longevity genes and aging interventions.

Modeling the human aging transcriptome across tissues, health status, and sex

“…Rf machine learning offers many unique advantages compared with other models, such as having a high predictive power, assigning a relative importance to different inputs, being non‐parametric, and having the capacity to automatically detect non‐linear relationships (Couronne et al, 2018; Touw et al, 2013). Rf was also recently used to generate accurate clocks in mice that can predict either age or life expectancy (Schultz et al, 2020).…”

“…Because aging clocks appear to routinely capture biological age—which correlates with various health parameters and outcomes—they have the potential to significantly accelerate anti‐aging intervention testing in animal models and anti‐aging clinical trials in humans. For example, the laboratory of David Sinclair recently used a machine learning model trained on frailty index components to predict the efficacy of anti‐aging interventions up to a year in advance in mice (Schultz et al, 2020). In humans, a small, exploratory study from the laboratory of Steve Horvath suggests that treatment with various pharmaceuticals, including metformin, can reverse biological age (Fahy et al, 2019).…”

“…The remaining four terms in the "healthy" cohort were extracel- up to a year in advance in mice (Schultz et al, 2020). In humans, a small, exploratory study from the laboratory of Steve Horvath suggests that treatment with various pharmaceuticals, including metformin, can reverse biological age (Fahy et al, 2019).…”

ARDD 2020: from aging mechanisms to interventions

“…The assessment of health state and life expectancy using deep learned clocks was further presented by Alice Kane, Harvard, USA. She developed a deep learned aging measure using a frailty index as a fast non-invasive mortality predictor for mice [22]. Another talk on estimation of chronological age was given by Anastasia Georgievskaya, Haut.AI, Tallinn, Estonia.…”

Pinpointing a Role for Vitamin D in Frailty: A Time for Animal Models?

“…In addition to us and others who model the Fried et al frailty physical phenotype [ 35 , 36 ], others have generated strategies to emulate the deficit accumulation model of frailty in mice [ 37 ]. Building from this latter strategy, Schultz et al have devised a method to estimate biologic age and life expectancy in mice [ 38 ], all potential areas for future studies involving vitamin D. Animal studies such as these also have the potential to inform clinical trials. Future clinical work may seek to investigate the benefits of higher levels of supplementation to prevent the onset of frailty as an example.…”