tAge: R package for transcriptomic clock application

tAge tAge R package provides API for Gladyshev Lab transcriptomic clock use. The package implements preprocessing pipelines, age prediction using Elastic Net and Bayesian Ridge models, and comprehensive visualization tools for transcriptomic age analysis. ⚠️ Please use the latest version of this record. This version (tAge 1.0.0) contains bugs that affect clock outputs: Mortality clocks were incorrectly rescaled by species maximum lifespan. Their output is a log10 hazard ratio and must not be multiplied by max lifespan; earlier mortality predictions are inflated by that factor (e.g. ×122.5 for human, ×48 for mouse). Relative (_scaleddiff / _yugenediff) clocks were not centred by default. When no reference group was specified, centring was skipped, adding a constant offset to every prediction. Both issues are fixed in version 1.1.0 and later, which reproduces the reference TACO application bit-for-bit. Predictions produced with 1.0.0 — especially for mortality — should be regenerated. ➡️ Get the corrected version via the “newer version” banner above, the concept DOI 10.5281/zenodo.19039289 (always latest), or GitHub: https://github.com/Gladyshev-Lab/tAge. Models availability The current tAge models can be obtained here: Transcriptomic clock models and rodent gene expression meta-dataset Citation Tyshkovskiy et al. (2026) Universal Transcriptomic Hallmarks of Mammalian Ageing and Mortality. Nature. License This package and the associated transcriptomic clock models are distributed under the MGB Open Access License 1.0. These materials may be used for non-commercial academic purposes, subject to the license terms. Commercial uses require a separate commercial license or agreement with Mass General Brigham.https://github.com/Gladyshev-Lab/tAge/blob/main/LICENSE For questions about commercial use, please contact MGB Innovation / Business Development & Licensing: innovation@partners.org licensing@partners.org