Authors release LivAge, an open tool that estimates mouse liver age from gene activity

Authors release LivAge, an open tool that estimates mouse liver age from gene activity

On August 21, Aging Cell published an article about LivAge. The tool takes RNA-seq results, which measure the activity of different genes in tissue, and estimates mouse liver age in months. The authors also released the calculation code.

In aging experiments, it is difficult to determine whether a diet, drug, or genetic modification changes the state of a tissue. The control and experimental groups may have the same chronological age, while differences in lifespan can take a long time to emerge. LivAge reduces a large table of gene activity measurements to a single value that researchers can use to compare groups.

The authors trained the model on 432 liver samples from healthy C57BL/6 mice collected across 23 studies. The animals ranged from one to 30 months of age. The training set included only control groups without genetic modifications or other interventions. The algorithm selected 268 genes whose combined activity determines the estimated liver age.

The final validation used 134 samples from four other studies that were not used to train or tune the model. The mean error was 1.53 months for males and 2.47 months for females.

The researchers then applied LivAge to separate datasets with known biological contrasts. The estimated liver age was higher in four mouse models of progeria, a group of diseases in which signs of aging appear earlier. It was lower after some genetic and dietary interventions that previous studies had linked to slower aging in mice.

On the LivAge website, a researcher uploads a table of raw RNA-seq results, and the service returns the estimated age in months along with a file containing the calculation. The open code and web tool give laboratories a consistent way to apply this calculation to new series of mouse liver experiments.